Textbook Group Processes Data Driven Computational Approaches 1St Edition Andrew Pilny Ebook All Chapter PDF

Group Processes Data Driven
Computational Approaches 1st Edition

Andrew Pilny
Visit to download the full and correct content document:
https://textbookfull.com/product/group-processes-data-driven-computational-approach
es-1st-edition-andrew-pilny/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Monitoring Multimode Continuous Processes: A Data-

Driven Approach Marcos Quiñones-Grueiro
https://textbookfull.com/product/monitoring-multimode-continuous-
processes-a-data-driven-approach-marcos-quinones-grueiro/
Computational Learning Approaches to Data Analytics in

Biomedical Applications 1st Edition Donald Wunsch
https://textbookfull.com/product/computational-learning-
approaches-to-data-analytics-in-biomedical-applications-1st-
edition-donald-wunsch/
Healthcare Big Data Analytics Computational

Optimization and Cohesive Approaches Intelligent
Biomedical Data Analysis 10 1st Edition Bhoi
https://textbookfull.com/product/healthcare-big-data-analytics-
computational-optimization-and-cohesive-approaches-intelligent-
biomedical-data-analysis-10-1st-edition-bhoi/
Data Driven Approaches for Healthcare-Machine learning

for Identifying High Utilizers 1st Edition Chengliang
Yang (Author)
https://textbookfull.com/product/data-driven-approaches-for-
healthcare-machine-learning-for-identifying-high-utilizers-1st-
edition-chengliang-yang-author/
Dynamic Modeling of Complex Industrial Processes Data
driven Methods and Application Research 1st Edition
Chao Shang (Auth.)
https://textbookfull.com/product/dynamic-modeling-of-complex-
industrial-processes-data-driven-methods-and-application-
research-1st-edition-chao-shang-auth/
Data Driven Treatment Response Assessment and Preterm

Perinatal and Paediatric Image Analysis Andrew
Melbourne
https://textbookfull.com/product/data-driven-treatment-response-
assessment-and-preterm-perinatal-and-paediatric-image-analysis-
andrew-melbourne/
Data Driven Fault Detection for Industrial Processes

Canonical Correlation Analysis and Projection Based
Methods 1st Edition Zhiwen Chen (Auth.)
https://textbookfull.com/product/data-driven-fault-detection-for-
industrial-processes-canonical-correlation-analysis-and-
projection-based-methods-1st-edition-zhiwen-chen-auth/
Statistical Process Monitoring Using Advanced Data-

Driven and Deep Learning Approaches: Theory and
Practical Applications 1st Edition Fouzi Harrou
https://textbookfull.com/product/statistical-process-monitoring-
using-advanced-data-driven-and-deep-learning-approaches-theory-
and-practical-applications-1st-edition-fouzi-harrou/
Group processes : dynamics within and between groups

Third Edition Brown
https://textbookfull.com/product/group-processes-dynamics-within-
and-between-groups-third-edition-brown/
Computational Social Sciences
Andrew Pilny
Marshall Scott Poole Editors
Group
Processes
Data-Driven Computational Approaches
A series of authored and edited monographs that utilize quantitative and computational
methods to model, analyze and interpret large-scale social phenomena. Titles within
the series contain methods and practices that test and develop theories of complex
social processes through bottom-up modeling of social interactions. Of particular
interest is the study of the co-evolution of modern communication technology and
social behavior and norms, in connection with emerging issues such as trust, risk,
security and privacy in novel socio-technical environments.
Computational Social Sciences is explicitly transdisciplinary: quantitative methods
from fields such as dynamical systems, artificial intelligence, network theory, agent
based modeling, and statistical mechanics are invoked and combined with state-of
the-art mining and analysis of large data sets to help us understand social agents, their
interactions on and offline, and the effect of these interactions at the macro level. Topics
include, but are not limited to social networks and media, dynamics of opinions, cul-
tures and conflicts, socio-technical co-evolution and social psychology. Computational
Social Sciences will also publish monographs and selected edited contributions from
specialized conferences and workshops specifically aimed at communicating new find-
ings to a large transdisciplinary audience. A fundamental goal of the series is to provide
a single forum within which commonalities and differences in the workings of this field
may be discerned, hence leading to deeper insight and understanding.
Series Editors
Elisa Bertino Larry Liebovitch
Purdue University, West Lafayette, Queens College, City University of
IN, USA New York, Flushing, NY, USA
Claudio Cioffi-Revilla Sorin A. Matei
George Mason University, Fairfax, Purdue University, West Lafayette,
VA, USA IN, USA
Jacob Foster Anton Nijholt
University of California, Los Angeles, University of Twente, Enschede,
CA, USA The Netherlands
Nigel Gilbert Andrzej Nowak
University of Surrey, Guildford, UK University of Warsaw, Warsaw, Poland
Jennifer Golbeck Robert Savit
University of Maryland, College Park, University of Michigan, Ann Arbor,
MD, USA MI, USA
Bruno Gonçalves Flaminio Squazzoni
New York University, New York, University of Brescia, Brescia, Italy
NY, USA Alessandro Vinciarelli
James A. Kitts University of Glasgow, Glasgow,
Columbia University, Amherst, MA, Scotland, UK
USA
More information about this series at http://www.springer.com/series/11784

Andrew Pilny • Marshall Scott Poole
Editors
Group Processes
Data-Driven Computational Approaches
Editors
Andrew Pilny Marshall Scott Poole
University of Kentucky University of Illinois
Lexington, KY, USA Urbana, IL, USA
ISSN 2509-9574 ISSN 2509-9582 (electronic)

ISBN 978-3-319-48940-7 ISBN 978-3-319-48941-4 (eBook)
DOI 10.1007/978-3-319-48941-4
Library of Congress Control Number: 2017930624
© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Contents
1 Introduction................................................................................................ 1
Andrew Pilny and Marshall Scott Poole
2 Response Surface Models to Analyze Nonlinear
Group Phenomena..................................................................................... 5
Andrew Pilny and Amanda R. Slone
3 Causal Inference Using Bayesian Networks............................................ 29
Iftekhar Ahmed, Jeffrey Proulx, and Andrew Pilny
4 A Relational Event Approach to Modeling Behavioral Dynamics........ 51
Carter T. Butts and Christopher Steven Marcum
5 Text Mining Tutorial.................................................................................. 93
Natalie J. Lambert
6 Sequential Synchronization Analysis....................................................... 119
Toshio Murase, Marshall Scott Poole, Raquel Asencio,
and Joseph McDonald
7 Group Analysis Using Machine Learning Techniques............................ 145
Ankit Sharma and Jaideep Srivastava
8 Simulation and Virtual Experimentation: Grounding
with Empirical Data................................................................................... 181
Deanna Kennedy and Sara McComb
v
Chapter 1
Introduction
Andrew Pilny and Marshall Scott Poole
For many young group researchers, learning about advanced statistical methods can
be quite the traumatic experience. Coupled with teaching, professional develop-
ment, and being theoretical experts in their domain, fine graining the ins and outs of
inferential statistics seemed like just another task on a full plate of work. Fortunately,
for many of us, there was a rescuer. In 2000, Andy Field published his first book,
Discovering Statistics Using SPPS for Windows, beginning a series of volumes ded-
icated to making statistics seem both easy and fun. Clarity was essential for Field,
whose volumes always provided relevant examples (usually very humorous), clear
screenshots, and example write-ups. Field’s volumes were vital for not only learn-
ing about statistics, but reducing anxiety and uncertainty the complexities of infer-
ential modeling.
However, the world has changed greatly since then, moving into what is gener-
ally referred to as the era of Big Data. Four characteristics generally characterize
Big Data (Gandomi & Haider, 2015): (1) volume (i.e., bigger size and magnitude),
(2) variety (i.e., more different types of data), (3) velocity (i.e., rate at which data is
created), and (4) complexity (i.e., complex data structures that require cleaning and
integration). But Big Data is not just about data per se, it is also about a new way
thinking about measurement (King, 2016). For instance, instead surveying groups
about their networks, we can now collect their interactions via their cell phones,
email, and social media (i.e., trace data). Unfortunately, one of the consequences of
Big Data is that many of the methods detailed by Field, which were exclusive vari-
ants of the general linear model, are either inappropriate or unsuited for much of the
data we have on groups today. For instance, for data on online groups (e.g.,
A. Pilny (*)
University of Kentucky, Lexington, KY, USA
e-mail: andy.pilny@uky.edu
M.S. Poole
University of Illinois, Urbana, IL, USA
e-mail: mspoole@illinois.edu
© Springer International Publishing AG 2017 1

A. Pilny, M.S. Poole (eds.), Group Processes, Computational Social Sciences,
DOI 10.1007/978-3-319-48941-4_1
2 A. Pilny and M.S. Poole
c oordination in Wikipedia), there can be millions of data points, which can results
in nearly every independent variable tested being statistically significant. Likewise,
interaction data from group members assumes a type of interdependence that vio-
lates many assumptions inherent in linear inference.
To address these issues, many researchers have called upon a paradigmatic
change in thinking, largely referred to as Computational Social Science (CSS)
(Cioffi-Revilla, 2013; Lazer et al., 2009). Computational social science is an inter-
disciplinary endeavor specifically tailored to handle the complexity of Big Data by
merging together social science problems with computer science methods. As
Wallach (2016) puts it, CSS can be thought of as research being undertaken by
groups of “social minded computer scientists and computationally minded social
scientists” (p. 317). The impact of CSS on group research has been especially nota-
ble. For instance, the new range of tools and thinking behind CSS has provoked
innovative ways of understanding different group dynamics (e.g., Klug & Bagrow,
2016; Shaw & Hill, 2014) and collecting group data (e.g. Madan, Cebrian, Moturu,
Farrahi, & Pentland, 2012; Radford et al., 2016).
Although the outlook of CSS is promising for the future of group research, there
is a looming problem (e.g., Alvarez, 2016): for all the new work being produced
using CSS methodology, there are few explicit avenues available to actually teach
these methods. In other words, pedagogy is has taken a back seat to publication. The
result is a sort of knowledge concentration or what boyd and Crawford (2012) refer
to as a digital divide between the small minority who have access to Big Data and
CSS resources and the majority who do not. Indeed, there are few graduate semi-
nars, workshops (often expensive if they do exist), or handbooks that make it easy
for the average social scientist to excel at CSS.
What is needed, therefore, is an “Andy Field book” for CSS, a resource to help
demystify these methods and make it accessible to anyone willing to follow the
white rabbit of CSS. To accomplish this goal, a resource would need to do several
things. First, it would need to emphasize a didactic, rather than an inquiry-laden
focus. That is, the primary objective is teaching rather than theory generation or
original contribution to research. Second, it would need to be transparent, which
means that codes and data should be shared and presented in a tutorial fashion.
Transparency is vital in an age where we see social science continuing to be criti-
cized for a lack of replication and secrecy regarding data and code. And finally, the
resource should be encouraging. The spirit behind such an endeavor should reflect
a growing notion that the more scholarly use of these methods, the better. As such,
opaque and ambiguous language, equations, and procedures should be avoided in
order to foster an environment that enables and empowers researchers to carry out a
similar analysis.
These three values represent the spirit behind this book. The authors were given
a relatively open format to write their chapters as long as it corresponded to a didac-
tic, transparent, and clear avenue for anyone to pick up and take off with. The diver-
sity of these chapters are quite evident: some are longer than others (e.g., Chap. 4:
Relational Event Modeling), some introduce needed theoretical introductions (e.g.,
Chap. 6: Social Sequence Analysis), some use computer code (e.g., Chap. 2:
1 Introduction 3
Response Surface Modeling), some use graphic interface programs (e.g., Chap. 5:
Text Mining; Chap. 3: Bayesian networks) and some may not even use data at all
(e.g., Chap. 8: Computational simulation).
Although no book on introducing CSS methods will be exhaustive, we aimed to
provide the audience with what might interest group researchers the most. For
instance, the growth of machine-learning is arguably one of the most dramatic
changes in inferential modeling during the last twenty years (Hindman, 2015).
Machine-learning models are often better equipped to handle Big Data because they
are not dramatically influenced by sample size, often do not make crude normality
assumptions, and have clear interpretations that explicitly acknowledge when the
model predicts both accurately and inaccurately. As such, we included two chapters
that explore different machine learning algorithms, Bayesian networks (Chap. 3)
and decision-trees (Chap. 7).
Likewise, there has been a renewed increase in group dynamics that openly
acknowledges time and order. In this case, group researchers can begin to seri-
ously consider dynamic rather than static notions of emergence (Kozlowski,
Chao, Grand, Braun, & Kuljanin, 2013). As such, Chap. 4 focuses on group inter-
actions by viewing networks as relational events (i.e., episodic interactions),
rather than relational states (i.e., enduring relationships). In this sense, relational
event modeling can reveal dominant patterns of interactions by predicting
ordered and even time-stamped histories of group interactions. Chapter 6 simi-
larly focuses on time and order, but highlights social sequences of activities. One
of the highlighted example of such a technique is that it can determine if group
members behave in a synchronized pace (i.e., entrainment), provoking an impor-
tant inquiry as to whether the emergence of group level properties are related to
group performance.
It also important to recognize the new types of data that can be exploited by
CSS methods. One example is the growing advent of analyzing text as data. In this
sense, Chap. 5 explores text mining procedures and the development of semantic
networks represented by co-occurrence relationships between different words and
concepts. Sometimes there is not enough data or something was missing from
data measurement. Chapter 9 deals with this through computational simulation
with empirical data. Finally, sometimes we have enough data on groups with
repeated observations that we can run quasi-field experiments. Chapter 2 adapts
response surface methodology, a common method in the natural and physical sci-
ences, to group research.
Lastly, as Alvarez (2016) notes, CSS is “developing at a dizzying pace” (p. 25).
While researchers are rapidly developing tools to provide unique and sometimes
ground-breaking insights into social inquiry, there is a need to pause and give back.
Many of the tools used by CSS researchers were not developed individually in a
vacuum. We owe a debt of gratitude to those who developed and taught us these
methods, and owe it to the next and current generation of CSS researchers to share
knowledge on how to use these methods. It can be seen as a sort of methodological
“pay-it-forward”. This book is one small attempt at such an endeavor.
4 A. Pilny and M.S. Poole
References
Alvarez, R. M. (2016). Computational social science: Discovery and prediction. Cambridge, MA:
Cambridge University Press.
Boyd, D., & Crawford, K. (2012). Critical questions for big data: Provocations for a cultural, tech-
nological, and scholarly phenomenon. Information, communication & society, 15(5), 662–679.
Cioffi-Revilla, C. (2013). Introduction to computational social science: Principles and applica-
tions. London: Springer.
Gandomi, A., & Haider, M. (2015). Beyond the hype: Big data concepts, methods, and analytics.
International Journal of Information Management, 35(2), 137–144.
Hindman, M. (2015). Building better models prediction, replication, and machine learning in the
social sciences. The ANNALS of the American Academy of Political and Social Science, 659(1),
48–62.
King, G. (2016). Preface: Big data is not about the data! In R. M. Alvarez (Ed.), Computational
social science: Discovery and prediction (pp. vii–vi1). Cambridge: Cambridge University
Press.
Klug, M., & Bagrow, J. P. (2016). Understanding the group dynamics and success of teams. Open
Science, 3(4), 1–11.
Kozlowski, S. W., Chao, G. T., Grand, J. A., Braun, M. T., & Kuljanin, G. (2013). Advancing
multilevel research design capturing the dynamics of emergence. Organizational Research
Methods, 16(4), 581–615.
Lazer, D., Pentland, A. S., Adamic, L., Aral, S., Barabasi, A. L., Brewer, D., … Gutmann, M.
(2009). Life in the network: The coming age of computational social science. Science,
323(5915), 721.
Madan, A., Cebrian, M., Moturu, S., Farrahi, K., & Pentland, A. (2012). Sensing the “health state”
of a community. IEEE Pervasive Computing, 11(4), 36–45.
Radford, J., Pilny, A., Reichelmann, A., Keegan, B., Foucault-Welles, B., Hoyde, J., et al. (2016).
Volunteer science: An online laboratory for experiments in social psychology. Social
Psychology Quarterly, 79(4), 376–396.
Shaw, A., & Hill, B. M. (2014). Laboratories of oligarchy? How the iron law extends to peer
production. Journal of Communication, 64(2), 215–238.
Wallach, H. (2016). Computational social science: Towards a collaborative future. In R. M. Alvarez
(Ed.), Computational social science: Discovery and prediction. (pp. 307–317). Cambridge
University Press.
Chapter 2
Response Surface Models to Analyze
Nonlinear Group Phenomena
Andrew Pilny and Amanda R. Slone
2.1 Introduction to Response Surface Methodology
Using Response Surface Methodology (RSM) is a lot like being a chef, mixing
together different combinations of ingredients to see which ones come together
to make the best dish. In this situation, strict linear thinking no longer applies.
For instance, adding just the right amount of salt to a dish can bring out the
sweetness in desserts or bump up the taste in more savory dishes. But, too much
salt can overwhelm the flavor of a dish, just as too little salt can leave it tasting
bland and unsatisfying. Chefs must find that perfect amount of salt that takes
their dish from acceptable to exceptional. In addition, chefs must consider how
the salt will interact with other ingredients in the dish. For example, salt inter-
acts with the yeast in bread to help create texture, and it helps sausage and other
processed meats come together by gelatinizing the proteins. Likewise, RSM
helps us find the optimal amount of an outcome variable based on two or more
independent variables.
This chapter will provide an introduction on how to use RSM to analyze nonlin-
ear group phenomenon. First, the chapter will outline a brief history and background
of the approach. Then, the chapter will walk the reader through a tutorial demon-
strating how to execute the second-order model using the PROC RSREG function
in SAS. Data previously collected from virtual groups in the game EverQuestII
(see Williams, Contractor, Poole, Srivastava, & Cai, 2011) will be provided as an
example.
A. Pilny (*) • A.R. Slone

University of Kentucky, Lexington, KY, USA
e-mail: andy.pilny@uky.edu; amanda.slone@uky.edu
© Springer International Publishing AG 2017 5

A. Pilny, M.S. Poole (eds.), Group Processes, Computational Social Sciences,
DOI 10.1007/978-3-319-48941-4_2
6 A. Pilny and A.R. Slone
2.2 Brief Background of RSM
Box and Wilson’s (1951) treatment on polynomial models provided the foundation
for RSM, which evolved and developed significantly (e.g., different variations of
designs) during the 1970s (Khuri, 2006). Like many statistical methods, RSM
developed in the natural sciences, but has yet to be applied extensively within the
social sciences given the amount of repeated observations needed for RSM. Indeed,
given the complexity involved in running controlled social science experiments and
the typically low rate of manipulations, though, it is no wonder that RSM has not
taken hold. However, with the advent of Big Data providing virtual Petri dishes of
human behavior, RSM has garnered new interest in the social sciences for its ability
to answer questions about complex group interactions (Williams et al. 2011. For
example, due to its emphasis on optimization (i.e., finding the right combination of
independent variables that maximizes a dependent variable), RSM has primarily
impacted the world of business and performance management.
2.3 Basic Processes Underlying RSM
RSM is a blend between least squares regression modeling and optimization methods.
More formally, RSM can be defined as the “collection of statistical and mathemati-
cal techniques useful for developing, improving, and optimizing processes” (Myers,
Montgomery, & Anderson-Cook, 2009, p. 1). Moreover, instead of trying to only
explain variance, RSM also seeks to clarify optimization. In other word, it is not
necessarily about how a set of independent variables explains a dependent variable,
but rather what combination of independent variables will yield the highest (or low-
est) response in a dependent variable. In order to do this, RSM requires at least three
variations in each variable, measured on a ratio or interval level.
To conduct an RSM test, there are typically five consecutive steps to go through
(SAS Institute, 2013): (1) the regression modeling, (2) lack of fit, (3) coding of
variables, (4) canonical analysis, and (5) ridge analysis. Each of these steps is
described in more detail below.
2.3.1 Step 1: Second-Order Regression Modeling
The most common and most useful RSM design is the second-order model because
it is flexible (i.e., not limited to linear trends), easy (i.e., simple to estimate using
least-squares), and practical (i.e., has been proven to solve real world problems;
Myers et al., 2009). The general linear model formula is identical to that which is
used when conducting a regression (Eq. 2.1):
2 Response Surface Models to Analyze Nonlinear Group Phenomena 7
y = f ( x1 , x2 ) + e (2.1)
In this general linear equation, y equals a response variable, x1 and x2 represent

predictor variables, and e equals the error term.
But, RSM uses the second-order model in order to fully determine the response
shape (i.e., the observed nonlinear trend). The equation for the second-order model
is as follows (Eq. 2.2):

y = b0 + x ¢ b + x ¢ B x (2.2)

In this second-order matrix equation, “b0, b, and B are the estimates of the inter-
cept, linear, and second-order coefficients” (Myers et al., 2009, p. 223) respectively.
One thing to note is that, unlike most other second-order regression models used
when conducting group research, the results provided in this model are preliminary.
That is, the results are used to determine linear, quadratic, and interactional relation-
ships between the independent variables, not to identify the response shape.
2.3.2 Step 2: Lack of Fit
Lack of fit is how well predicted repeated observations match the observed data. In
other words, lack of fit of the second-order model indicates that the predicted values
of the data do not look like the observed values (see Montgomery, 2005, p. 421–
422). For example, though salt (independent variable) may be shown to influence
taste (dependent variable) in a second-order model (i.e., statistically significant),
when we compare the predicted responses to actual taste ratings (e.g., feedback
from customers), there are major discrepancies. This indicates a poorly fitting
model.
When we have more than one observation on an independent variable, there are
several things to look out for when calculating lack of fit. First, it is important to
differentiate pure error from lack of fit error. Pure error is more common in regres-
sion modeling is determined by looking at the sum of squares (Eq. 2.3) variability
between each repeated observation of the independent variables (yij) and the average
value of the response variable ( y i):
m ni
SSPE åå ( yij - yi )
2
(2.3)
i =1 j =1
Lack of fit error is different because it uses a weighted version of yij and looks at
the actual observed value of the dependent variable, not the average. The Equation
(2.4) can be calculated by taking the sum of the difference of the average value of
the response variable ( y i) the fitted value of the response variable ( y i), and weight-
ing it by the number of observations at value of the independent variable (ni):
m 2
æ ö
SSLOF åni ç yi - y i ÷ (2.4)
i =1 è ø
From there, an F-test (Eq. 2.5) can be derived using mean squares (MS) from
both equations to determine whether or not a quadratic model is even necessary to
replace a reduced first-order model. For instance, if the lack of fit test is not signifi-
cant for a first-order model, then there could be a reasonable argument that a second-
order model is not event needed:
MSLOF
F0 = (2.5)
MSPE
Likewise, if the test is statistically significant for a second-order model, then by

Occam’s Razor (i.e., law of parsimony), we have evidence that a quadratic model
might not be appropriate.
2.3.3 Step 3: Coding of Variables
Despite requiring variables to be measured at the interval or ratio level, RSM does
not simply examine multiple sets of linear relationships. Instead, RSM conducts an
experiment of sorts, and organizes variables into conditions to see which results in
the optimal output. As such, to make it easier to conduct the canonical analysis
(Step 4) and ridge analysis (Step 5), recoding values is a convenient way to examine
the response shape at multiple values of the independent variables. As Lenth (2009)
put it, “Using a coding method that makes all coded variables in the experiment vary
over the same range is a way of giving each predictor an equal share in potentially
determining the steepest-ascent path” (p. 3). In addition to simplifying the calcula-
tion, recoding the variables also produces results with respect to the original values
of the independent variables. A common way to recode variables, as in the SAS
package, is to do the following (Eq. 2.6):
Original value - M
Coded value = (2.6)
S
whereas “M is the average of the highest and lowest values for the variable in the
design and S is half their difference” (SAS Institutive, 2013, p. 7323). For instance,
if there were five observations of on salt, ranging from two ounces to ten ounces,
then the data for salt are stored in coded form using the following (Eq. 2.7):
Salt value - (10 - 2 )

xsalt = (2.7)
4
2.3.4 Step 4: Canonical Analysis of the Response System
The next step is to conduct a canonical analysis of each of the conditions. The pur-
pose of the canonical analysis is to determine the overall shape of the data. For a
first-order model, this is typically done through a method of steepest ascent or
descent, wherein a linear shape determines which region of values creates an opti-
mal response. However, for a second-order model, the shape can look more three-
dimensional given the addition of interaction and polynomial terms. Here, we go
back to our original Eq. (2.2) of a second-order response in matrix form (see Myers
et al., 2009, p. 223):

y = b0 + x ¢ b + x¢ B x
To optimize the response ( y ) and locate the stationary point (xs) (i.e., the point of
highest response in the dependent variable) we can set the derivative of y equal to 0:
¶ y
= b+2Bx = 0 (2.8)
¶x
and then solve for the stationary point:
1 -1
xs = - B b
2
In these equations, b equals a vector of first-order beta coefficients:
és ù
êb1 ú
ê . ú
ê ú
ê . ú (2.9)
ê . ú
ê ú
ê b ú
ë qû

And B includes quadratic (diagonals) and interaction (off-diagonals) beta
coefficients:
é ù
ê b 11, b 12 / 2, ¼, b 1q , / 2 ú
ê ú
ê b 22, ¼, b 2 q , / 2 ú (2.10)
ê ú
ê ú
ê ú
ë sym. b qq , / 2 û
For instance, consider if we trying to maximize taste ( y ) with salt (x1) pepper
(x2). After running a clean second-order model (i.e., no lack of fit), we find that:
é 3.65 ù é -1.22 -0.25 ù

b=ê ú , B = ê -0.25 -2.66 ú
ë 4.69 û ë û
then
1 -1
xs = - B b
2
1 é -0.84 0.08 ù é 3.65 ù
=- ê
2 ë 0.08 -0.38 úû êë 4.69 úû
é1.34 ù
=ê ú
ë0.75û
To compute this equation using matrix algebra, the following R code can be
used:
B = matrix ( c ( 1.22, - .25, - .25, - 2.66 ) , nrow = 2, ncol = 2, byrow = TRUE )

b = matrix ( c ( 3.65,4.69 ) , nrow = 2, ncol = 1, byrow = TRUE )
(
x = -.5 * ( solve ( B) ) %*%b )
x
As such, the predicted stationary point for taste based on salt and pepper (x1, x2)
is 1.34 and 0.75. If the hypothetical fitted second-order model is
y = 69.65 + 12.22 x1 + 3.45 x2 - 9.33 x12 - 6.32 x22 - 4.65 x1 x2
then the predicted highest response of taste æç y s ö÷ would be 63.63 by plugging in the
è ø
optimal values for salt and pepper. It can then be re-expressed in the canonical sec-
ond order form (this will be useful for later, see Montgomery, 2005, p. 446):
y = ys + l1 w12 + l2 w22
= 63.63 + 1.34 w12 + 0.75w22
where w1 and w2 are canonical variables (i.e., latent variables in relationship with
the original independent variables).
From this point, it is necessary to determine the shape of the stationary point. The
eigenvalues (λ) of the canonical analysis give indication to the nature of the shape
(see Montogmery, 2005, p. 446):
| B - lI |= 0 (2.11)
If the eigenvalues for each independent variable are negative, then a maximum
stationary point has been reached. A maximum stationary point looks like a hill,
meaning that there is a point that indicates a high response. For most research, this
is good news because it means that some combinations of variables entered in the
model to produce a maximum response in the dependent variable. On the other
hand, if they are all positive, then this indicates a minimum stationary point, mean-
ing that the data will look like a valley. For most research, this is bad news because
it means that some combinations of variables entered in the model to produce a
minimum response in the dependent variable, unless a decrease in the dependent
variable was what was desired of course.
Finally, if the eigenvalues are mixed, this indicates a saddle point, meaning that
maximum or minimum solutions are not found, but rather multiple regions of high
and low variables exist. In other words, the data will look like a series of hills and
valleys, or perhaps even a plateau. For instance, a high value of x1 and low value of
x2 may produce the highest value of y, while at the same time, a low value of x1 and
low value of x2 may also produce the same value in y. Moreover, if they are all very
close, or are at zero, then there is a flat area, meaning that there was little to no rela-
tionship between the independent variables and the response variable. Beyond look-
ing at the eigenvalues, a two-dimensional contour plot is also a visual that can easily
determine the shape of the response surface.
From our current example,
|B − λI| = 0
é -1.22 - l -0.25 ù
ê -0.25 =0
ë -2.66 - l úû
By taking the determinant of the matrix:
( -1.22 - l ) ( -2.66 - l ) - ( -0.25 * -0.25) = 0

l 2 + 3.88l + 3.183 = 0
The solution, using basic completing the square calculus, is λ1 = −1.177 and
λ2 = −2.70. As such, because both eigenvalues were negative, it indicates a maximum
stationary point. This means that the canonical values for salt and pepper would yield
the highest value of taste based on the data the researcher has collected.
2.3.5 Step 5: Conduct Ridge Analysis if Needed
Often when a saddle point is found, or if the researcher wants additional informa-
tion regarding a maximum or minimum point, a ridge analysis can be performed.
The purpose of a ridge analysis is to “anchor the stationary point inside the
experimental region” and to give “some candidate locations for suggested improved
operating conditions” (Myers et al., 2009, p. 236). In other words, the ridge analysis
provides an estimated response value of y for each of the different values in the
independent variables.
For instance, consider if the eigenvalue for pepper was essentially zero, but salt,
as we found out, was significantly less than zero (see, Montgomery, 2005, p. 447).
From this point, we would want to see what values of salt would yield a high amount
of taste by analyzing the predicted response in taste from different values in salt. If
the example formula was
y = 69.65 - 14.87 x1 + 7.94 x2 - 0.33 x12 - 8.89 x22 + 13.65 x1 x2
and the resulting response in canonical form was

y = ys - 13.561 w12 - 0.02 w22
then we know we can pay more attention to salt because a single unit in the w2
canonical variable would results in a 13.56 unit change rather than a small 0.02 unit
change moved in the w2 direction. In Table 2.1, a ridge analysis used this informa-
tion to produce a line of predicted values that might indicate a trend:
Table 2.1 Example ridge analysis table

Order Estimated response in taste Un-coded value for salt Un-coded value for pepper
1 65.24 2 1
2 59.17 3 0.9
3 49.53 4 0.8
4 36.32 5 0.7
5 19.55 6 0.6
From here, one can see how the decreasing levels of salt are related to a higher esti-
mated response in taste, which could prove useful for future design of experiments.
2.4 RSM in Context
To demonstrate the usefulness of RSM in group research, this exemplar study

employs data gathered from a download of data on 100,000 characters over 5 months
in the Massive Multiplayer Online Game (MMOG) Everquest II (EQII).
2.4.1 About the Game
Commercially launched in November 2004, this game was estimated to have about
200,000 active subscribers in North America alone as of early 2008, the year in
which the data was drawn from.1 These players participate in thousands of teams
over the 5 months, making it possible to draw much larger samples and making it
possible to identify large samples of teams. Moreover, they incorporate precise met-
rics for the for team performance outcomes. As such, a random sample of 154
unique groups (i.e., no shared members) was analyzed for this tutorial.
As in most MMOGs, EQII players create a character and advance that character
through challenges in a social milieu, typically banding together with other players
for help and companionship. For each character, a class is chosen to fit some varia-
tion of the three basic archetypes found in nearly every fantasy MMO: damage-
dealer, damage-taker and damage-healer. Each archetypal role has different
capabilities, weaknesses and strengths, and the choice of class then determines how
players develop their characters and how they will interact in the game environment
and with other players. Players can communicate with others in the game through
text messaging and voice chat.
Following a loose storyline, players use their characters to complete various
tasks (quests) in order to earn virtual items such as currency and equipment. One
important performance metric is number of “experience points” gained during a
quest. Players must accumulate experience points to advance their character level.
The character level is a fundamental indicator of players’ success in the game. It not
only represents a quantitative measure of players’ skill and competence, but also
determines whether players have access to certain quests and other game content,
locations, and equipment. Until they attain the maximum level of 70, the accumula-
tion of experience points is the only way for players to increase their character level.
The amount of experience points associated with a given quest is associated with the
difficulty of the quests and the value of the items won. Therefore, experience points
can be used as a simple yet powerful indicator of players’ performance at the com-
mon tasks in the game.
At the opposite end of the spectrum, a player can die during a quest. When a
player dies in the game, they are not gone forever, but do pay a cost. For instance,
for several minutes, the character is very vulnerable and cannot use many of their
capabilities until they have had time to refresh many of their spells, buffs, and item
effects. Moreover, their armor takes a significant amount of damage and if com-
pletely destroyed, the character will have to find a shop to get new armor or get it
repaired. Finally, unless they are revived from a teammate, they will likely revive at
a location far away from where the quest was being performed. As such, it is in the
team’s interest to avoid death because it can hinder their progress in the quest.
1
There is no definitive evidence for the exact size of the population on Everquest II. The number
200,00 is estimated from multiple professional and fan sites such as http://www.mmogchart.com
and http://gamespot.com
This study focused on group (i.e., heroic) quests and teams of three to six mem-
bers. Generally, groups undertaking heroic quests include characters with different
capabilities and skills. As discussed earlier, experience level is an important indica-
tor of players’ capabilities and competence, and groups often have members with
different experience levels. The diversity in experience levels in a group can influ-
ence team processes substantially (Valenti & Rockett, 2008). Groups also typically
are composed of members with different archetypal roles (i.e., damage-dealers,
damage-takers, and damage-healers).
The groups in EQII that undertake heroic quests resemble the action teams
described by Sundstrom, De Meuse, and Futrell (1990) in that they have short-term
projects with clear goals and standards for evaluation, and members take on specific
highly-interdependent roles. Their projects are the quests in the game, which require
players to complete certain activities, such as finding objects or information, or kill-
ing a monster. Success or failure is clearly indicated by whether the quest is com-
pleted or not and whether or not members are killed during the quest. Analogous
real world teams include military units, emergency medical response teams, and
surgical teams.
2.5 Dependent Variable
2.5.1 Team Performance
Team performance was measured using two metrics. The first was the amount of
experience points each player earned during the quest. These were obtained through
the back-end database. Throughout the quest, characters earn points for successfully
completing required tasks (i.e., defeating a monster, finding hidden objects).
Likewise, death was the second and separate indicator of team performance. The
total amount of deaths was calculated and the lower the number of group deaths, the
better the performance.
2.6 Independent Variables
2.6.1 Complexity
Task complexity scores for each group were obtained through individually cod-
ing each quest. Detailed descriptions of each quest were obtained through ZAM
EverQuest II, the largest EQII online information database. ZAM also features
EQII wikis, strategy guides, forums, and chat rooms. Graduate and undergradu-
ate researchers independently coded each quest based on the general definition of
task complexity given by Wood (1986). According to Wood (1986), complexity

entails three aspects: (1) component complexity (i.e., the number of acts and
information cues in the quests), (2) coordinative complexity (i.e., the type and
number of relationships among acts and cues), and (3) dynamic complexity (i.e.,
the changes in acts and cues, and the relationships among them). These features
were used to code the complexity of each quest (Mean = 21.11, SD = 17.35,
Min = 4, Max = 81).
2.6.2 Difficulty
Difficulty scores for each quest were obtained through Sony Online Entertainment.
Each quest is given a static difficulty score ranging from 1 (least difficult) to 70
(most difficult). To create a variable that most closely resembled how difficult it was
for the group attempting it, we subtracted the difficulty of the level of the quest from
the highest player’s character level. Thus, a negative number indicates that the group
has at least one player that has a character level much higher that the quest they are
attempting, meaning that it will likely be quite easy. On the other hand, a positive
number indicates that everybody in the group has a character level below the quest
difficulty level, meaning that it will likely be quite difficult to complete
(Mean = –2.52, SD = 6.76, Min = –31, Max = 13).
2.7 Control Variables
2.7.1 Group Size
The more group members, the more likely there are opportunities for groups to both
earn experience points and die. As such, to account for group size, we used group
size as a covariate. Groups ranged from three (67.3 %), to four (22.2 %), five
(5.2 %), and six (5.2 %) members. Since most groups has three members, the group
size of three was used a reference point.
2.8 Data Analysis
The current example carries out RSM in SAS, through the proc. rsreg procedure.
SAS is used here because it has perhaps the simplest code, though other programs
can easily implement RSM like R and JMP.
2.8.1 Controlling for Group Size
Another benefit of using SAS is that the procedure, including contour plots and
ridge analysis, are all done through specifying a few lines of code:
data rsm;
set rsm;
g6 = ( groupsize = 6 ) ;
g 4 = ( groupsize = 4 ) ;
In the above line of code, the first thing that we must do is create the covariate
variable. Since we want qualitative variable for each group size, we create four dif-
ferent variables and call them g6, g5, g4, and g3.
2.8.2 Experience Points: A Minimum Stationary Point
The next line of code runs the RSM procedure:
ods graphics ;
proc rsreg data = rsm plots = ( ridge surface ) ;
model experience _ pts = g6 g 4 g5 g3 difficulty complexity / covar = 4 lackfit;
ridge max min;
run;
( odsgraphicsoff; )
The first line (ods graphics on;), simply tells SAS to turn on the ODS Statistical
Graphics (Rodriguez, 2011). These graphics are necessary to produce the contour
plots that show the predicted response based on different values of the independent
variables. The second line of code does two things. First, it specified the data, which
we have named “rsm” (proc rsreg data = rsm). Second, it tells the program which
types of plots we want form the output. In this case, we want a ridge and surface plot
(plots = (ridge surface)). The third line of code specifies the model variables. In
model one, we are analyzing experience points as a function of quest difficulty and
complexity while treating group size as a covariate. When reading this line of code,
the dependent variable should come directly after the model term followed by an
equal sign (model experience_pts=). The independent variables should come next
(g6 g4 g5 g3 Difficulty Complexity), making sure to have the covariates come first.
The covariate command lets the program know that the first four variables are to be
treated as covariates and not included in the canonical and ridge analysis (covar = 4).
The final line of the model command is the lack of fit test, telling the command
to include it in the output (lackfit;). Finally, we want to include the ridge analysis to
find values of the independent variables that predict a maximum or minimum
response in experience points (ridge max min;). After these commands are properly
arranged, we must tell the program to run it (run;). Turning the ODS Graphics off
is useful because it might make future commands run a bit slower, even if they are
not using the ODS Graphics.
2.9 Results
The following figures contain screenshots from the actual SAS output to ease in initial
interpretation. Figure 2.1 contains the results from the least squares regression, includ-
ing the interaction and polynomial terms. Before the results, however, are some
descriptive information, including how the two independent variables were re-coded
The SAS System
The RSREG Procedure
Coding Coefficients for the

Independent Variables
Factor Subtracted off Divided by

difficulty -9.000000 22.000000
complexity 42.500000 38.500000
Response Surface for Variable experience_pts:

experience_pts
Response Mean 3232.605656
Root MSE 1359.462094
R-Square 0.1533
Coefficient of Variation 42.0547
Regression DF Type I Sum of Squares R-Square F Value Pr > F

Covariates 3 10950853 0.0348 1.98 0.1204
Linear 2 23611455 0.0751 6.39 0.0022
Quadratic 2 9855992 0.0314 2.67 0.0729
Crossproduct 1 3758993 0.0120 2.03 0.1560
Total Model 8 48177294 0.1533 3.26 0.0019
Fig. 2.1 Coding coefficients and ANOVA (experience points)

Residual DF Sum of Squares Mean Square F Value Pr > F
Lack of Fit 120 235646022 1963717 1.55 0.1084

Pure Error 24 30485732 1270239
Total Error 144 266131755 1848137
Standard Parameter Estimate

Parameter DF Estimate Error t Value Pr > |t| from Coded Data
Intercept 1 3661.765063 361.579354 10.13 <.0001 2326.341716
difficulty 1 -24.793149 29.353031 -0.84 0.3997 -587.744282
complexity 1 -53.178658 25.183288 -2.11 0.0364 -792.042760
difficulty*difficulty 1 3.050587 1.336332 2.28 0.0239 1476.484094

complexity*difficulty 1 1.246778 0.874219 1.43 0.1560 1056.020971
complexity*complexity 1 0.515613 0.342138 1.51 0.1340 764.267643
g6 1 1088.103762 514.943437 2.11 0.0363 1088.103762
g4 1 354.285863 276.326599 1.28 0.2019 354.285863
g5 1 292.070086 511.576337 0.57 0.5689 292.070086
g3 0 0 - - - 0
Factor DF Sum of Squares Mean Square F Value Pr > F Label
difficulty 3 13954485 4651495 2.52 0.0606 difficulty
complexity 3 27768937 9256312 5.01 0.0025 complexity
Fig. 2.2 Model coefficients and lack of fit (experience points)
for the canonical and ridge analysis, and descriptives for the dependent variable,
which in this case is experience points (M = 3232.61). The omnibus analysis of vari-
ance table compares the different models (e.g., linear, quadratic, cross-product, covari-
ate) to an intercept-only model in order to determine how much of an effect they add.
For instance, because the quadratic terms by themselves (F = 2.67, p = 0.07), more or
less, provide a better fit than an intercept-only model, it means they will likely be
influential predicting an optimal or minimal response surface.
Figure 2.2 displays information on the lack of fit test and individual estimates for
each independent variable. Overall, the lack of fit test was just above a 0.10 t hreshold
for significance (p = 0.11). While this is generally acceptable as a rule of thumb, it
points to some concern about how well the model predicted the actual response of
experience points. Nevertheless, there were both linear and nonlinear effects in the
model. For instance, there was a negative linear relationship with complexity
(t = −2.11, p = 0.03), meaning that groups earned more experience points with less
complex tasks. On the other hand, while there was not a linear relationship with
difficulty (t = −0.84, p = 0.40), there was a quadratic effect (t = 2.28, p = 0.02),
Fig. 2.3 Canonical

analysis (experience
points)
meaning that there is a certain difficulty peak where groups tend to earn more expe-
rience points. To further investigate that point, a canonical analysis is useful here.
Figure 2.3 shows the results of the canonical analysis. Because the eigenvalues
for difficulty (λ1 = 1757.25) and complexity (λ2 = 483.50) were both positive, this
means the unique solution is a minimum. In other words, a unique combination of
difficulty and class can yield a solution in which groups earned the least amount of
experience points. As such, these two variables cannot tell us much about high per-
forming groups, but do tell us a lot about low performing groups. Moreover, because
difficulty was over three times the value of complexity, it means that experience
points changes more rapidly along changes in difficulty compared to complexity.
Finally, the table below gives the solution for the predicted minimum stationary
point of 2271.66 experience points at a value of −8.59 for difficulty and 61.96 for
complexity. Because the mean values for each variable is −2.52 and 21.11, this
means that groups perform the worst when they choose quests that are about 40 units
higher in complexity than average and when the groups highest member is about
8 units less than the quest value, which is higher than average.
Figure 2.4 shows the ridge analysis for a minimum solution. The quadratic effect
for difficulty is clearly evidence here as the values fluctuate from moving higher from
−9 to −7.87, then decreasing from −7.87 to −13.07. This is important because the
relationship as demonstrated by the regression model is not linear, suggesting that the
difficulty of the quest compared to the highest-level character in the group has a tip-
ping point (~−8.59). On the other hand, though there is an overall negative linear
trend with complexity according to the regression model, the ridge solution paints a
more complicated picture. For instance, almost equal predicted responses are
obtained with a complexity value of 42.50 and 80.33. These results are in line with
Fig. 2.4 Ridge analysis of minimum response (experience points)
the critical value threshold demonstrating that a complexity value near 60 is where
groups are predicted to perform the least, much higher than average (Mcompelxity = 21.11).
Although no maximum solution was found, a ridge analysis for maximum ascent
tends to demonstrate simple linear effects for both difficulty and complexity (see
Fig. 2.5). More specifically, groups are predicted to perform better when the charac-
ter level of the highest character approaches the same level of the quest and the
complexity of the quest increases. This makes sense because it means that the quest
should not be as challenging for the group if they have at least one character in the
group that is close to the quest difficulty level. The quest is still complex enough for
group members to do activities that will give them a chance to earn points. However,
no solid conclusions should be drawn from this. Instead, it may serve as an impetus
to collect more data for future analysis.
Finally, Fig. 2.6 is a visualization of the response surface analysis as a contour
plot, with covariates fixed at their average values. This means that this plot is most
relevant for groups of three, which were the majority of groups playing this game.
The minimum solution can be easily visualized by looking at the large ring r epresenting
values below 3000. Values closer to the center of that ring are the lowest predicted
values of experience points. If you cross the intersection between the two critical
values of −8.59 for difficulty and 61.96 for complexity, one can pinpoint to the center
of the ring. The circles represent the predicted values for each observation.
Fig. 2.5 Ridge analysis maximum response (experience points)
2.9.1 Model for Deaths: A Saddle Point
For the model with deaths as the response surface, we use the same code except
switch the dependent variable form experience points to deaths:
ods graphics on;

proc rsreg data = rsm plots = ( ridge surface ) ;
model deaths = g6 g 4 g5 g3 difficulty complexity / covar = 4 lackfit;
ridge max min;
run;
odsgraphics off;
The initial outputs in Fig. 2.7 details similar information about the coded vari-
ables and analysis of variance.
As you can see in Fig. 2.7, there is a significant difference between an intercept
only model and the linear, quadratic and cross-product models, suggesting that the
Response Contour for experience_pts with Design Points

5000
4000 4000
10
1500
0
3000
Standard Error
difficulty
-10 1000
-20
4000
5000 500
6000
-30
7000
0 20 40 60 80
complexity
Fixed at: g6=0.0523, g4=0.2222, g5=0.0523, g3=0.6732
Fig. 2.6 Contour plot (experience points)
variables have considerable influence on deaths. However, as demonstrated by

Fig. 2.8, the full model has a significant lack of fit, meaning that the average val-
ues of death deviate more than we would expect by chance from the predicted
responses of deaths.
Indeed, although there are significant effects regarding the difficulty term
(t = 3.45, p < .01) and overall interaction term (i.e., difficulty*complexity, t = 2.38,
p = .02), the lack of fit finding puts a hitch into the entire analysis because it means
that we cannot generalize much of the subsequent canonical and ridge analysis.
From here, this usually means the researcher might look into some additional rea-
sons for the lack of fit. For instance, there may not be enough variability in deaths
and it might be useful to transform it to make it look more normally distributed (e.g.,
log linear transformation). Alternatively, the researchers might attempt to add more
data or additional explanatory variables. Nevertheless, for demonstration, we will
carry on with the canonical and ridge analysis.
As expected, there was no unique solution because of the saddle point response
shape as demonstrated by the mixed signs of the eigenvalues (see Fig. 2.9).
Nevertheless, the eigenvalue for difficulty (λ1 = 12.51) is quite larger for complexity
(λ2 = −1.03), suggesting that there was more variability regarding changes in
difficulty. Because there was a significant quadratic interaction, it is useful to look
The SAS System
The RSREG Procedure
Coding Coefficients for the Independent

Variables
Factor Subtracted off Divided by

difficulty -9.000000 22.000000
complexity 42.500000 38.500000
Response Surface for Variable deaths: deaths
Response Mean 3.718954

Root MSE 6.680704
R.Square 0.1725
Coefficient of Variation 179.6393
Regression DF Type I Sum of Squares R-Square F Value Pr > F

Covariates 3 322.150612 0.0415 2.41 0.0698
Linear 2 394.317475 0.0508 4.42 0.0137
Quadratic 2 371.536805 0.0478 4.16 0.0175
Crossproduct 1 251.930606 0.0324 5.64 0.0188
Total Model 8 1339.935498 0.1725 3.75 0.0005
Fig. 2.7 Coding coefficients and ANOVA (deaths)
at a maximum ridge analysis to see what exact levels of difficulty were more associ-
ated with more deaths.
The ridge analysis complicates things even further because although the
regression model suggested a nonlinear effect on difficulty, the ridge analysis
does suggest a linear relationship (see Fig. 2.10). In other words, the more groups
attempt quests that have difficulty levels higher than their highest level character,
they are more likely to die in that attempt. Again, however, this might be due to
a lack of fit.
Finally, the contour plot in Fig. 2.11 visually demonstrates the relationship
between difficulty and complexity as it relates to the number of deaths incurred on
a question. The wide open space in the middle indicates the least amount of deaths,
but does not reveal a solution because those groups varied too widely on complexity
and difficulty. Moreover, the bottom left and top right corners specify very high
predicted values of deaths, meaning that no maximum solution could be found
either because the existence of these high values occurs at seemingly opposite ends
of the spectrum. That is, a high number of deaths can occur at a combination of
either high complexity and low difficulty, or high difficulty and low complexity.
Another random document with
no related content on Scribd:
IV.
ROUBEKLAG.
Gedurende die dae wat volg, beskryf die koerante, tot in alle
besonderhede, die treurspel wat daar vér in die woestyn afgespeel
is. Aan die roemryke nagedagtenis van die heldhaftige kaptein Kibert
word alle eer bewys. Sy portret en lewensbeskrywing kom in die
koerante. Sy te korte loopbaan word deur alle harte betreur. In hul
eensaamheid, op Maupas, oorstelp van droefheid, ontvang die
moeder en dogter in stille gelatenheid die talryke betuiginge van
simpatie, wat uit die hele land na hulle toestroom, van die Regering,
van die kommando-maats van Marcel, van bekende en onbekende
mense. Hulle steun mekaar, om hul verlies beter te kan verstaan en
te kan dra; en hul enigste troos vind hul in gebed en in hul
wedersydse liefde vir mekaar.
Meteens kom nou die deftige wêreld tot besluit om die voortou te
vat in die uiting van algemene simpatie—die deftige wêreld, wat eers
die rug toegekeer het aan die famielie Kibert in hul eerlike verlies
van hul fortuin.
Mevrou Delourens voel dat sy iets moet doen. Sy haal ou
jongjuffrou Sonjon oor—as honorêre presidente van die Rooi-
Kruisvereniging—om voor te gaan in die instelling van ’n lykdiens,
wat met groot ophef sal plaasvind in die groot kerk van Chamberie.
Dis naamlik van belang om die nagedagtenis van die roemryke dode
in genadige beskerming te neem, en om deur al die deftigheid uit te
blink bo sy burgerlike afkoms. Die outoriteite sal ook uitgenooi word
om teenwoordig te wees; hulle sal aan die diens iets vernaams
meedeel—en hul afwesigheid sal die koerante van die politieke
teenpartye ’n stok in die hand gee! In elk geval sal dit dus goed
gaan.
Alles is nou mooi agtermekaar—alles en almal bymekaargesoek,
die diens bepaal, die uitnodiginge uitgestuur; mejuffrou Sonjon en
mevrou Delourens word offisiëel afgevaardig na Maupas om nou net
nog die famielie Kibert se toestemming tot die kerkdiens te vra.
Mevrou Marthenay (Alida) gaan saam met haar ma; sy wil haar
roubeklag doen—wat sy innig diep voel!—aan mevrou Kibert en aan
Paula, en sy het nie gedurf om alleen te gaan nie.
Dis die eerste dae van Maartmaand. Die sneeu smelt op die
somber, modderige velde, op die verspoelde paaie. Onder
laaghangende, swaar wolke, omring van nakende, swart,
verwaarloos uitsiende bome, lê die ou buitehuis daar, droefgeestig
en verlate.
—Ek sou my tog nie, jaar-in en jaar-uit, op so ’n plek lewendig wil
begrawe nie, sê mevrou Delourens aan mejuffrou Sonjon terwyl die
rytuig die verlate boomlaning opry.
—Die kerk is te vèr van hulle af, sê die ou jongjuffrou, met ’n
vroom gesig.
Die ou siel het nog nooit uitgevind, dat God oral is nie, en dat sy
nie die hele land hoef rond te reis—veral na die plekke waar sy dit
lekker het—om Hom te vind nie.
Ou Marie, die bediende, sien die deftige rytuig, en waag dit nie om
toegang te weier aan die hoë dames nie, alhoewel sy strenge orders
het om niemand toe te laat nie. So gou as haar ou bene haar kan
dra, gaan sy sê dat daar mense is. Mevrou Kibert is verdrietig
daaroor en sê:
—Maar ek het jou tog gesê ek wil niemand sien nie. Sy draai na
Paula: Ek het nie die moed om mense te ontvang nie. Waarom kom
mevrou Delourens ons hinder? Ons het niks aan haar nie, wat wil sy
hê?
—Ek weet tog nie ma, sê Paula; sy staan op, klaar om uit die
kamer te gaan.
—Help my om hulle te ontvang, Paula.
Mevrou Kibert kyk haar dogter aan, en sien hoe haar bleek gesig
beef en ’n besliste trek aanneem, aantonende wat daar in haar
omgaan.
—Paula, smeek sy, verlaat my nie; ek is bangerig en onhandig,
dink daaraan. As ’n mens kwaad gedoen het, is dit gouer vergeet as
wanneer ’n mens goed gedoen het. As sy met my oor die verlede
praat, sal ek nie weet wat om haar te antwoord nie. Bly by my,
Paula.
Paula aarsel nie meer nie en gee ’n teken aan Marie om die
mense binne te laat, terwyl sy sê:
—Ek bly hier, ma.
Mejuffrou Sonjon is maar min gewend aan pligpleginge, en laat die
woord aan mevrou Delourens:
—Wat ’n wrede slag vir u! sê mevrou Delourens, terwyl sy nader
kom na mevrou Kibert—die ou vrou moet aan die skoorsteenmantel
gryp om te kan opstaan van haar stoel; dan groet mevrou Delourens
ook vir Paula, terwyl sy voel hoe die meisie haar met ’n ferme en
vyandige blik aankyk. Sy wens dat Paula uit die kamer is.
—Ja, antwoord Marcel se moeder, God beproef ons.
So gee sy dadelik ’n ernstige en godsdienstige wending aan die
gesprek. Mejuffrou Sonjon slaan haar oë op na die hemel, net of sy
alleen die vereiste outorisasie het om die goddelike tussenkoms in te
roep:
—Hoe algemeen is tog die beklag wat u in u rou ontvang! sê
mevrou Delourens weer. Die eenstemmige betuiginge van
bewondering vir die heldhaftigheid van die kommandant, die
algemene simpatie en rou . . . . In ons dae van volksheerskappy
word verdienste nie meer genoeg geëer nie. Dikwels moet die dood
eers die volle waarde daaraan toeken; en eers na ’n onherstelbare
verlies, betreur ’n mens dit bitter dat jy te laat van die verdienste
kennis geneem het.
Mevrou Kibert is bewoë sodra van haar kind gepraat word. Sy dink
by haarselwe: Nou wil sy ekskuus maak omdat sy Marcel afgestoot
het; eindelik sien sy haar fout in en het daar spyt van; maar mevrou
Marthenay moet nie saamgekom het nie: haar teenwoordigheid is
pynlik vir ons. Die ou vrou kyk mevrou Delourens aan, en die
opheldering in haar smart-verwoeste gelaat lyk soos ’n sonstraal in
’n blaarlose bos in die winter. Maar Paula is nie so maklik te
oorrompel nie, hoewel sy daar nie aan kan twyfel of mevrou
Delourens heeltemal sonder bygedagte so praat nie.
Nà ’n oomblik stilte, begin mevrou Delourens nou met die doel van
haar besoek.
—U sal dit dus heeltemal natuurlik vind dat ons die begeerte voel
om eer te bewys aan so ’n waardevolle nagedagtenis. Die hele land
neem deel in u rou, maar veral die mense van aansien, tot wie u
seun behoor het deur sy famielie en sy verdienste.
Sy skep ’n slaggie asem, en nou sy voel dat sy mooi op stryk raak,
swaai sy haar oog snel oor haar toehoorders. Mejuffrou Sonjon
staan goedkeurende buiginkies te maak met haar lanklysige,
uitgedroogde lyf. Alida, in haarself gesluit, kyk na die bedroefde
aangesigte van mevrou Kibert en van Paula—haar ou skoolmaat. Sy
voel so sleg, dat sy haar hand op haar bors moet druk: die snikke
wat sy inhou, verstik haar. Sy sou haar hele hart uitstort voor die
moeder en dogter, maar sy durf nie. Sy probeer om saggies die hand
te vat van Paula, wat naas haar sit. Maar beslis trek Paula haar arm
weg: sy het nog niks vergeet nie.
Die kragtige stem van mevrou Delourens weerklink weer deur die
stille voorkamer:
—Al die vernaamste dames van die Rooi-Kruis en van die hoogste
kringe het meteens ’n aansporing gevoel om te vra dat daar ’n
plegtige lykdiens gehou sal word in die groot kerk van Chamberie. Al
die predikante uit die omtrek het beloof om te kom. Die Regering sal
goed verteenwoordig word. Die diens sal so groot en pragtig wees,
glo my, mevrou, dat dit die nagedagtenis van die grote dode waardig
sal wees.
Sonder ’n woord te praat, het mevrou Kibert haar aangehoor; sy
antwoord eenvoudig:
—Ek dank u, mevrou; en bedank asseblief ook vir my al die
dames vir hul goeie voornemens. Ons het, na ons eie beurs, ’n klein
diens laat hou. Al was dit koud en vèr, ons vriende het almal gekom.
Die generaal van die kommando was self ook daar, en ’n groot
klomp offisiere. Baie dankie, maar ons het nie behoefte aan nog
meer publieke plegtighede nie.
—Ja, mevrou, ek begryp u goed. ’n Famielie hou nie daarvan dat
vreemdelinge indring by hul rou nie. Maar dis hier ’n besondere
geval. Die dood van kommandant Kibert is ’n publieke ramp. Die
verlies van u seun is ’n verlies vir die hele land. Sy lewe en sy dood
is die land tot eer. Dit moet u dus nie verwonder as ons hom groot
dank betoon nie. Dit spreek vanself dat die geldmiddele van ’n
famielie maar beperk is. Maar laat ons dit in hande neem. Ontsê ons
nie die plesier nie . . . .
Dadelik slaan sy die ongelukkige losgelate woord dood:
—Die droewige plesier wat ’n lykdiens-gebed ’n mens laat voel.
Die plegtigheid is ’n gebed vir ons. U, wat so ’n brawe kristin is—kan
u die gebede weier wat ons wil aanbied? Kan u dit oor u hart kry om
ons deelneming in u rou van die hand te wys?
—Die kerk stel belang in die plegtighede en oefeninge, sê
mejuffrou Sonjon—haar godsdiens is praalsugtig en aristokraties.
Alida het aan die muur ’n vergrote portret ontdek van Marcel; en
vir niks het sy nou nog oog nie as vir die fiere jonkman, wat sy so
sonder moed liefgehad het.
Onderwyl aarsel mevrou Kibert nie omtrent haar antwoord nie,
maar omtrent goeie en beleefde woorde tot inkleding van haar
antwoord. Mevrou Delourens kom haar hier ’n verbetering aanbied
op haar eie beskeie lykdiensie, wat so sonder praal of pronk of glans
was, en nou aangevul moet word deur ’n minder armsalige, deur ’n
skitterender en wêreldse vertoning. Dis die rykdom wat die armoede
besoek en genadig onder beskerming wil neem. So neem Paula dit
op; sy kook inwendig en kyk haar ma skerp aan met haar donker oë,
waar lig in flikker—soos blitse in die nag. Maar mevrou Kibert sien in
die aanvraag niks meer nie as ’n eerbiedige hulde aan haar kind; en
hoewel sy so’n mededinging—wat sy as onnuttig beskou—van die
hand wil wys, probeer sy elke woordjie te vermy wat die minste
wrywing sou kan veroorsaak.
Paula vrees dat haar ma te beskroomd is, en het ’n halwe gedagte
dat sy besluiteloos is—dus waag sy dit om haar voor te spring:
—U plan het ons diep getref, mevrou, en ons stel dit na verdienste
op prys; dit spyt ons om so’n guns af te wys. Die nagedagtenis van
my broer is al behoorlik geëer. Ons verlang nie na meer publieke
roubetuiginge nie as ons reeds ontvang het. God meet Sy seëninge
nie af na die maat van die mooiste plegtighede nie.
Mevrou Delourens maak of sy volstrek geen waarde heg aan wat
Paula te sê het nie: sy draai na mevrou Kibert. Die ou vrou begryp
die beweging en sê:
—Ja, mevrou, dis soos Paula sê.
Mejuffrou Sonjon stik amper, en slaan haar oë hemelwaarts, terwyl
die meesteres van Chenée—ongewoon om dwars getrek te word—
die aanval hernuwe:
—Ek kan u weiering nie verklaar nie. Ons vra niks meer nie as om
u ons simpatie op die mees natuurlike manier te betoon. Al die
dames, mejuffrou Sonjon, gravin Lavernay, barones Ambelard, is dit
eens met my. Ek verteenwoordig hulle by u. Ons predikant sal ook
sy bes doen.
Sy dink om indruk op die ou vrou te maak deur al die hoë adel op
te noem. Sy weet nie, en sy kan nie weet, tot wat ’n graad van
onverskilligheid die lewe mevrou Kibert gebring het ten opsigte van
wêreldse sake en mense nie.
Paula kry jammer vir haar ma in haar verleëntheid. Sy kap nou om
daar ’n end aan te maak.
—In Chamberie was dit bekendgemaak dat ons ’n lykdiens sou
hou. Al ons vriende het gekom, selfs van verre plekke; daar het selfs
onbekende mense gekom om ons te bekla. Maar dis my vertel dat u
bank leeg was, mevrou; ek wou dit nie glo nie.
Aan die set voeg sy nog toe:
—As my oudste broer, wat nou hoof van die famielie is, dink dat
verdere pligpleginge nog bepaald nodig is, dan sal hy ons dit wel laat
weet. Ons sal na sy begeerte handel. My moeder en hy is alleen
bevoeg om hierin iets te doen.
Mevrou Delourens sien nou dat alle moeite verniet is; sy staan op:
—Dit spyt my, mevrou, dat ons die misverstand nie kan wegneem
nie. Ek het soiets nie hier verwag nie. Maar ek sien dat u dogter
groot invloed oor u het.
—Ek is dit heeltemal eens met haar, sê die ou vrou, terwyl ook sy
met moeite opstaan.
Sy keur goed wat Paula gesê het, maar sy sou gewens het dat dit
alles op minder oorheersende toon gesê was. Sy vrees dat haar
besoekers aanstoot sal neem, en dit pynig haar, en ’n kleurtjie vlam
op in haar bleek wange. Terwyl sy die dames uitgeleide doen, merk
mevrou Delourens die ligte gloed op. Sy soek na wraak, en sy dink
dit daarin te vind: met katterige spotsug gee sy die ou vrou ’n steek:
Dag, mevrou. Maar kyk, u sien daar waarlik goed uit! Mooi so! Dit
verbaas ons en ons is daar bly om.
Die ou vrou haar oë skiet vol trane: sy is nog maar alte gevoelig vir
onregvêrdige beoordeling. Oud, krom en gebroke, soos sy daar is,
moet sy elkeen met medelye vervul, behalwe die teleurgestelde vrou
van die wêreld. Soetjies sê sy, terwyl die bloed weer uit haar wange
wyk:
—Ek hoop die Heer sal my gesond hou! My taak is nog nie op ’n
end nie.
Sy dink by dié woord aan Paula; haar onsekere toekoms maak die
ou vrou ongerus en laat haar nog vashou aan die lewe. Maar die
voorkamerdeur is agter hulle toegemaak. Sy volg die dames tot by
hul rytuig, hul gaan sit, maar sien dat Alida agtergebly het.
—Ek sal haar roep, sê mevrou Kibert, en sukkel die treedjies weer
op.
Alida het alleen agtergebly by Paula en laat eindelik die vrye loop
aan haar verdriet:
—Paula, my liewe Paula, wil jy my toelaat om jou ’n soen te gee?
O, ek het ook so baie geween. Kan jy ooit weet? Ek is so bedroef
vandat . . . . . vandat hy nie meer leef nie. Ag, jy kan dit nie verstaan
nie!
Paula staan onbeweeglik, ongenaakbaar; met verwondering kyk
sy die elegante jonge vrou aan, met die rein en skoon gelaatstrekke,
wat haar om ’n guns vra; en sy dink aan wat verby is:
—Wat sal dit help? sê sy.
En alhoewel sy sien dat Alida haar oë hol is en haar gelaat asvaal,
sê sy nog daarby, tussen haar tande:
—Is dit nie jou skuld, min of meer, dat ons vandag so ongelukkig is
nie?
Hierdie swak, verliefde kind het geweier om met haar broer te trou,
en daaraan skryf sy dit toe dat hy so na die dood begin te verlang
het; nadat hy die laaste keer van Chenée teruggekom het, het sy dit
baiemaal opgemerk, in woorde wat hy hom so kalm laat ontval het
oor die dood. Sy, wat nou hier by haar staan te huil, het toe net nodig
gehad om één woordjie te praat om weer lewenslus op te wek in die
vurig beminnende hart van haar broer en daarby ook die vertroue
wat selfs die gevaar in bedwang hou. Dié woordjie het sy toe nie
uitgespreek nie, hoewel sy hom liefgehad het. Was sy onverskillig
gewees vir sy liefde, dan sou sy nou nie skuldig wees nie; maar haar
lafhartigheid was sterker gewees as haar liefde.
Alida staan nog maar te snik:
—Ag, ek is nog ongelukkiger as jy.
Dit lyk nou darem vir Paula of Alida haar droefheid regtig
diepgevoeld is. Bewoë sluit sy haar ou skoolmaat in haar arms; en—
soos in die ou dae by hul grappies—so meng die twee nou hul trane
in verdriet.
—Ek het hom liefgehad, fluister Alida.
—Waarom wou jy dan nie toestem nie?
Sy ontboesem haar nou eindelik heeltemal, en so tussen haar
snikke deur sê sy:
—Jy, Paula, kan jou trane ’n vrye loop gee. Maar ek moet my
vrolik hou terwyl ek die dood in my hart dra . . . . My liewe Paula,
mag God jou behoed van ooit te moet dra wat ek nou te dra het! . . .
. . En dis my eie skuld, Paula . . . . Ag, dis waar, ek sou vandag
liewer sy weduwee wil wees as wat ek nou is.
So kom Paula dan te wete hoe haar vriendin doodgemartel word
deur innerlike smart. So op die oog sou jy gesê het dat Alida
gelukkig was: die gedoente van die groot wêreld laat hom skaars
hoor tot by Maupas. En kyk, nou sien sy meteens die onmiddellike
en altyddurende straf wat opgelê word aan dié wat bang is vir die
lewe.
Alida leun op Paula haar skouer, net of sy smeek om hulp.
Alhoewel sy haar klere aanhet van pels, beef sy oor haar hele
liggaam. Paula soen haar, sy neem haar betraande gesig tussen
haar hande:
—Alida, ek het tog so jammer vir jou. Maar jy moet moed hou; jy
moet vergeet. Dink aan jou kindjie. Maak van haar ’n dapper vrou.
—Ek het hom liefgehad, herhaal sy nog ’n keer saggies.
Mevrou Kibert kom nou binne. Sy sien hoe die twee mekaar
omhels; en sy voel waarom hulle ween.
—U ma wag vir u, mevrou.
Sy soek na nog ’n woord en sê saggies:
—Dankie vir die besoek.
Dis ’n woord van vergifnis. Alida neem die ou vrou haar hand en
soen dit. Sy droog haar oë af, kyk nog vir die laaste keer na Marcel
se portret en loop haastig die kamer uit . . . .
Die rytuig gaan deur die blaarlose boomlaning en uit by die hek.
Mevrou Delourens was ongerus oor die lank wegbly van haar
dogter; sy kyk haar nou aan, met besorgdheid, met gehegtheid en
jaloersheid. Sy vermy sorgvuldig om haar te bekla oor die weiering
van mevrou Kibert en oor die houding van Paula. En terwyl die rytuig
by die eikebos uitry, lê sy haar hand op Alida haar arm en sê
soetjies, terwyl mejuffrou Sonjon sit en rondkyk:
—Nou sien jy hoe verstandig jou moeder was.
Die jong vrou kyk haar vraend aan. Sy vervolg:
—Natuurlik. Het ek jou toegelaat om met kommandant Kibert te
trou, dan was jy vandag weduwee.
Alida antwoord nie. Met skrik ontdek sy dat haar siel in baie
opsigte nog heeltemal onontgin is; en sy vra haarself of sy as
weduwee nie minder ongelukkig sou wees nie. Die smart wat die
noodlot ons oplê, is dieper, maar minder neerdrukkend as dié
waaraan ons self skuld het, deur ons swakheid, deur ons
lafhartigheid vir die lewe. Die eerste soort breek ons harte, maar
reinig en versterk ons daarna. Die laaste soort verslyt ons sonder
enige voordeel na te laat, en laat ons langsamerhand laer en laer
buk onder slae wat niks veredelends meebring nie. Het sy die beste
van die twee gekies? . . . . Die heldhaftige dood te beween van ’n
man wat sy bo almal sou verkies—dit lyk vir haar nog ’n aangenamer
lot as haar hele lewe lank die veragtelikheid te moet beween van dié
man wat altyd haar lewensmaat bly . . . .
V.
JAN.
Die famielies wat in die verte deur die ongeluk getref word, wat
sou hul nie gee om deur iemand wat ooggetuie was van die bloedige
offer, te hoor vertel van hul liewe dode nie, van die besonderhede
wat so kort en bondig ontwyfelbare waarhede laat voel in verband
met die treurige gebeurtenis, al sou sulke besonderhede ook hul
wonde weer oopmaak en hul trane weer laat vloei. Ja, hul is gelukkig
selfs in hul ellende, as hul maar die noukeurige waarheid kan hoor,
opdat die dood minder duister bly, en die onsekerhede weggeneem
word wat by dag ’n kwelling is vir hul gees en snags verander in
benoude drome . . .
Verskeie maande is al verloop na die veldslag in die Sahara. Van
die twee vrouens in rou op Maupas, het die één nog krommer
geword en op haar bleek gelaat het die glimlaggie vir altyd
versterwe; die ander een het regop en fier gebly, maar haar jeug
versmadend, gee sy haar bitter en hopeloos oor aan haar dae wat
verbygaan. Omsluit deur eensaamheid en stilte, gaan hul nie meer
dorptoe nie, en stap geen deur meer in nie as van die armoedige
huisies waar altyd na hul uitgesien word.
Nietemin, as hul die stap van die brieweman hoor op die sand van
die agterplaas, dan gee hul harte nog ’n ruk. Die goeie ou man skep
vreugde in die belangryke rol wat hy speel, hy laat hul nie wag nie,
op die posseëls kykend roep hy hul al van vèr toe: „Hier is één van
Parys . . . van China . . . . van Algiers.”
—Dankie, Ravet; Marie is in die kombuis, laat sy vir jou iets te
drinke gee.
In die briewe lê nou die hele plesier wat die huis nog ken. Daar
kom nou meer as anders. Uit die verre lande wedywer die kinders
van mevrou Kibert met mekaar om haar droewige ou dag met hul
liefde te verlig. Die briewe uit Afrika is van Jan Berlier; hulle vertel
net van Marcel en van sy roemryke dood. In sy laaste brief skryf Jan
dat hy teen die end van Meimaand huis-toe kom. Dis nou Meimaand:
alles is éne sonlig en blomme. Met verlange kyk hul al uit op Maupas
of daar nog nie iemand aankom in die verlate boomlaning nie, waar
die kastaiingbome trots hul wit bloeiselpluime in die lug dra . . . .
Die jonkman wat stadig teen die bult opstap, onder die bome wat
om die ou huis groei, is nie meer die galante vryer van Isabella
Orlandi nie. Hy het darem nog altyd sy mooi skraal postuur en sy
ligte dog vaste stap. Maar sy gebruinde gelaat het iets manlikers, en
dit lyk of sy oë elke ding meer besonderlik raaksien. Sy sorglose
jonkheid het oorgegaan in ’n leeftyd van nadink en wilskrag.
Hy het die aand vantevore aangekom. Vroeg vanmôre het hy die
rosevilla van sy oom verlaat. Hy asem die lug in van sy ontwakende
geboortegrond, van die herlewende natuur. Die veld lyk nog kouerig,
en is hier en daar oordek met violet en purper dampe, soos ’n
jongvrou wat langsaam haar ooglede opslaat, haar uitrek, en haar
sluierbedekking afskuiwe. So verras Jan die ontwakende lente, die
lewensvreugde wat met die daglig begin. Hy bewonder die tere
groen wat in Meimaand die velde en bome oorwaas, en hy ruik die
lug van die sagte jong blaartjies in die heinings. Hy soek na die
welbekende kerktorings teen die hang—die een byna in die groen
bome begrawe, die ander teen ’n bos van kastaiingbome, soos ’n
grysaard wat die koelte soek. Selfs die kranse van die berge het in
die môrelig hul dreiende aansien verloor. Onder die heldere lug
glimlag die hele natuur, en skyn vertroulik haar bevalligheid en
betowering aan te bied, belofte te gee van vrug en oes.
Jan draai hom om en sien in die verte, soos ’n plaat pêrelmoer en
goud, die meertjie lê, slapende nog in die môreson. Onder die
strelende sonstrale begin die water wellustig te rimpel. Die jonkman
stap aan. Teen die bult staan die villa Chenée, met oop rame in die
môrewindjie. Hy dink met plesier terug aan sy vyf-en-twintig-jarige
leeftyd en aan Isabella met haar mooi rooi lippe—net so klaar om te
soen as om te babbel. Hy dink terug en sê verwonderd:
—Amper vier jaar het ek hierdie plek, en vir haar, nie gesien nie.
Dit lyk my baie langer. Ek was toe maar ’n kalfie, wat met die lewe
gespeel het. Maar hy dink nie lank aan haar nie. In die eikebos hoor
hy ’n meisie sing wat die skape oppas: dis ’n ou liefdeliedjie:
Bo-op die berg

sing die voëltjies so.
Ek neem my roer
met my mee na bo.
Vier wou ek skiet,
maar weggevloë
is almal nou,
en geen één het ek, o!
Die stem klink so helder soos fonteinwater. In ’n draai van die pad
sien hy die skape, agter hulle die meisie, donker teen die lug, in die
omlysting van die bome wat oor die pad in mekaar groei. Dis ’n
dogter van vyftien of sestien jaar so gesond en fris as net ’n mooi
boerenoointjie kan wees. Sy sing weer verder:
My liefie se hartjie
het ek geraak.
O, het ek my liefie
nou seergemaak?
Sy stap verby Jan; hy glimlag.

—Dag, meneer Jan, sê sy, en buig met haar hoof.
—Ken jy my dan? vra hy, verbaas.
—Maar seker. Ek is die dogter van die boer op Maupas.
—Jeannette?
—Ja, om u te dien.
—Maar jy was so hoog soos my kapstewel, en nou is jy langer as
die koringhalme.
Daar is niks wat ons so laat voel hoe die tyd verbygaan nie as die
kinders wat groot word en wat ons maar min sien. Die meisie voel
gevlei en begin te lag, en haar vreugde werk aansteeklik. En terwyl
sy verder aanstap, sing sy weer die laaste versie:
My liefie se hartjie
het ek geraak.
O, het ek my liefie
nou seergemaak?
En tot by die jonkman, wat nog altyd onbeweeglik bly staan by die
eikebome, dra die windjie die laaste, flou-gehoorde woorde aan:
’n Bietjie maar net—

maar dit maak my dood!
O, soen my net éénkeer—
en weg is die nood.
Hy slaan sy oë in die rondte op die vorme van die natuur, wat hom
omring, op die jong boomblare, op die golwende gras van die
weilande, op die jongmeisie, wat haar jeug dra soos die oorlopende
wynbeker. Hy asem die geure in van die veld en die môrelug van die
bosse. En die lug van die ou tuisplek maak hom weer vol lewenslus.
Nou dat hy ondervind het dat skone dae verbygaande is—nou wil
hy daar ten volle van geniet. Jongmense stel nie waarde op hul
bestaan nie as hul gedagteloos hulle plesier, ligsinnigheid en
verstrooiing najaag, en alles wat die tyd maar versnel en doodmaak
gelyk. Dis lewensgevaar, dis hartstog, dis liefdesmart of die aanblik
van die dood, dis diepe droefheid wat hulle meteens tot stilstand
bring voor die ontmaskerde gesig van die lewe—soos ’n mens aan
die end van ’n tuinpad meteens ’n koue grafsteen sien onder rose
verberg—dié wat die nag nie ken nie, sal hy met dieselfde gretigheid
as ons die bedreigde prag van lig geniet en die vorme wat weer in
die donker verlore sal gaan?
Op die hoogtepunt van sy jonkheid aangeland, begryp Jan dit nou
beter. ’n Ander, meer donkerblou hemel, ’n ander dor en kaal land
het sy gevoeligheid geskerp en volmaak gemaak. En veral het
ander, droewiger aandoeninge sy hart by herhaling getref, net soos
die beitel van die beeldhouer ’n klip tref en die onnodige stukke laat
wegspat om ’n beeld te laat ontstaan. ’n Gevoel van dankbaarheid
laat hom die volle en opwekkende gevoelens wat hierdie
voorjaarsmôre hom gee, in verband bring met die bloedige môre toe
sy vriend geval het. So ’n doodval van sy aanvoerder ná die
oorwinning, so’n deurboorde voorhoof waarop die verstand sy setel
gehad het, so’n verkilde hart wat vol liefde was, soveel werkkrag en
moed, afgebreek soos ’n boom in bloei—kyk, dis dit wat ons wys op
die nietigheid van die mens en as teenstelling—die dag meer lig gee!
Daar, voor die bleek gesig van Marcel, uitgestrek op die grond,
skoon, ernstig en vreedsaam in sy onbeweeglikheid—’n skoonheid,
so stil tevrede, so kalm, dat hy dit nooit sal vergeet nie, so roerend in
die omgewing en in die omstandighede—daar het hy die lus gevoel
om te lewe, met sy hele hart en sonder vrees, en die behoefte om
die ewige werk van die dood teen te spreek.
Die ou hek van Maupas staan oop, soos altyd. Jan loop op in die
kastaiinglaning. Hy asem die geur van die bloeisels in. Hy weet dat
daar in ’n oomblik weer trane sal vloei, pynlik maar weldadig, en dat
hy die droewige verdienste sal hê van ou wonde weer oop te maak.
Die gruis kraak onder sy voete.
Mevrou Kibert sit op stoep te werk—met stram vingers—in die
koel môrelug. Sy staan op en kyk wie die besoeker is. Sy sien die
jonkman.
—Jan, is dit jy? Ek het tog so na jou verlang!
Met die eerste oogopslag sien hy die spore van haar beproewing
op haar. Sy het krommer geword, en gryser. Maar met verwondering
herken hy ook op haar vermaerde gesig die vreedsame uitdrukking
van vroeër.
—Mevrou, o, mevrou!
Haastig storm hy die treedjies op, en, net of dit natuurlik so kom,
buig hy oor haar en gee haar ’n soen.
Sy probeer vergeefs om nie te huil nie en fluister die naam van
Marcel.
—Kom binne, sê sy. In die voorkamer kan ons beter van hom
praat.
Sy loop sukkelend vooruit. Sy maak ’n deur oop en roep:
—Paula, hier is Jan Berlier.
—Ek het gisteraand gekom, sê die jonkman. En ek kom al vroeg
vanmôre omdat ek u graag gou wil sien.
—Dis baie goed van jou. Ek het wel geweet dat jy gou sou kom.
Verskeie dae al het ons oë die pad opgepas.
Paula kom binne en gee Jan haar hand. Haar mooi swart hare laat
haar mat-bleek gelaatskleur uitkom. Daar is geen vuur meer in haar
donker oë nie. Nog meer fier en regop as vantevore, dra sy trots
haar verbryselde hart. Alhoewel Jan heeltemal besig gehou word
deur die treurige storie wat hy vertel, verbaas dit hom om in die jong,
ernstige gelaat en in die stywe houding so’n lewensgeringagting te
sien, Sy is verbaas om ’n verandering in hom te ontdek: met die jare
het hy ’n beslister en fermer houding aangeneem—iets soos Marcel.
En in die klein voorkamer van die buitehuis—waar ’n sonstraal
deur die geslote luike kom—roep die getuie van sy dood die held
weer op uit die harde Afrikaanse grond waar hy rus, waar hy vir sy
land geval het, en bring hom terug in sy famieliekring. Jonk lyk hy,
groot, skraal en gespierd, sy hoof omhoog, gebiedend, begaaf met
die meerderheid wat sy verskyning laat voel, die aanvoerdersgees,
die gewilde kalmte wat ’n hoofman uitwys.
Op die piano sien Jan sy roos-omkranste portret.
Hy praat van hom soos hy dit sou verlang het: op eenvoudige en
waardige wyse. Hy het ’n mooi, waar manier van vertel. Sy woord
verjaag alle swakheid en wanhoop, hy bemoedig en vertroos en
weet selfs die dood voor te stel as voltrekking van ’n bestemming.
Die twee vrouens het geween by sy aankoms, maar aangenaam is
sy verhaal nou vir hul harte.
Hy het sy maat nie sien val nie. Dit was net dagbreek toe die
geweerskote hom laat opspring het; hy het sy manskappe dadelik
bymekaargeroep. Die kommandant het, ’n entjie daarvandaan, die
vyand dadelik aangeval. Die sersant, wat by hom was, het my vertel
hoe hy gesterwe het. Ek was op die linkervleuel, hy in die middel. Hy
het hulle uit hul posiesie geskiet en toe agternagesit. Teen die rooi
môrelug het ek sy swart skadubeeld gesien. Die sersant het hom
gewys op ’n duikie in die grond, waar miskien nog van die vyand kon
skuil. Hy gee ’n stap vooruit, voel meteens aan sy voorhoof, bly nog
’n sekonde regop staan, en val toe met ’n slag teen die grond.
Mevrou Kibert bêre haar gesig in haar hande, en daar is trane in
Paula haar oë, alhoewel sy probeer om haar te bedwing.
—Hy het nie meer geroer nie, sê die kaptein verder—hy het geen
pyn te ly gehad nie. Die dood het hom getref in die setel van sy
gedagte: reg voor sy voorhoof. En hy het aan sy vaderland gedink,
en aan julle.
—En aan God, nie waar nie? fluister die arme moeder.
—Ja, aan God. Ek moes bevel voer in sy plek. Maar hy het die
hele oorwinning al behaal. Toe ek hom eindelik kon soek, het hulle
hom ’n entjie opsy gedra, onder ’n palmboom. Vergeefs het ek my
oor hom heen gebuig. Ons dokter het my droefgeestig aangekyk—
hy het hom al ondersoek. Ons samelewe het broers van ons
gemaak. Ek het hom liefgehad soos julle hom liefhet. En daar het ek
hom beween soos julle, in jul naam. En ek het gesien—dié droewige
vreugde mog julle nie smaak nie—ek het die kalme, stille vrede
gesien wat hy in die dood gehad het. Dit het hom ’n uitdrukking
gegee van ewige skoonheid. As ek hom so sien in my herinnering
dan kry ek net gedagtes aan sieleadel en krag. Die aand vantevore
het hy nog met my meegegaan na my slaapplek. Dit was ’n heldere
sterrenag. Ons het baie oor die ou tuis gepraat. Hy het gepraat oor u
en oor Paula, hy het julle kort vantevore nog gesien. Hy het geen
droewige voorgevoel gehad nie, maar vir die dood het hy geen vrees
gehad nie. In sy baadjiesak het hy hierdie brief gehad, wat ek u
gebring het—dit het sy laaste hartslae gevoel.
Mevrou Kibert herken haar eie handskrif; sy kyk op—haar gelaat
vol moedersmart. As sy haar woorde kan kry, sê sy:
—En nou, nou rus hy in die vrede van God. Jan, sê my, waar is hy
begrawe?
—Op die slagveld, mevrou, tussen twee ander offisiere van laere
rang; aan sy voete lê die gesneuwelde soldate.
Paula sê:
—Ons het inligtinge gevra omtrent wat daar gedaan moet word dat
hy oorgebring kan word na Chamberie. Hy sal hier rus in ons
famielie-grafkelder, by my vader, wat, net soos hy, op die veld van
eer neergeslae is, by my sustertjie Thérèse.
Jan kyk haar aan. Hy weet dat die famielie Kibert maar arm is.
Vertroostend, oortuiend, byna gebiedend wil hy hul die kosbare en
nuttelose plan afraai:
—Waarom hom terug te haal? Die plek waar hy lê, dra die naam
van sy oorwinning, daar het hy getriomfeer. Watter graf sou hom
beter pas? Sou hy ooit ’n beter monument kon verlang het?
—Hulle sal hom daar gou vergeet.
—Nee, Paula, dis tog nie so nie. Al die grafte het opskrifte en word
goed versorg. Op syne staan sy naam, sy rang en die drie
oorwinninge wat sy lewe kenmerk.
Paula kyk weg en dink na, en Jan se oë is op haar fyn sy-gesig.
Nog altyd trou aan haar broer, sê sy weer:
—Ons sou hom so graag digteby ons wil voel, om op die steen te
kan kniel wat hom bedek.
—Julle besit tog sy nagedagtenis, sy gedagte. Wat van Marcel
oorbly, is tog iets anders as sy stoflike oorskot.
—Ja, sê mevrou Kibert, dinkend aan sy onsterflike siel.
Die suster van Marcel dring nie verder aan nie. Maar Jan merk dat
die trane oor haar wange rol.
—Hy was ons trots en my lewe, snik sy; en met sagter stem laat
sy daarop volg: Hy het al lank ’n voorgevoel gehad van sy lot.
—Dit was die Heer Se wil, sê haar moeder. Ons ken Sy weë nie.
Somtyds lyk Hy vir ons so wreed, dat ons lus het om te murmureer.
Nogtans is Sy goedheid grensloos.
Jan is bewoë. Hy neem in sy hand die bewende, gekerfde ou
hand, en, vol eerbied, soos Marcel ook eendag gedaan het, druk hy
daar ’n soen op. Hy staan op en, die twee vroue aansiende, spreek
hy met ernstige stem nog hierdie laaste lofrede uit op die dode—
miskien sal dit nog iets uitwerk op Paula, wat nog die minste getroos
en die mismoedigste is:
—Ja, sy kort loopbaan was kompleet. Deur sy wil, sy moed, het hy
dit tot ’n blywende voorbeeld gemaak. Moet ons hom nie liewer beny
in plaas van hom te bekla nie? Sonder vrees het hy die dood
tegemoet getree. Weet ons waar of wanneer die dood ons afwag?
Hy kan ons elke oomblik ontmoet. Dit alleen is van belang: sonder
vrees en sonder ophou steeds al ons kragte in te span tot bereiking
van ons bestemming. Ons moet die dode eer, dog ook vertroue
behou in die lewe.
Met haar donker oë, wat weer opvlam, kyk Paula die jonkman aan;
hy gee haar nuwe moed. Is hy dan nie meer die ligsinnige, jong
offisier wat gekheid maak met flirterige meisies nie? As sy terugdink,
dan vind sy dat sy hom altyd behandel het met ’n soort van
geringagtende verdraagsaamheid, miskien was dit ’n vae gevoel van
ongeopenbaarde hartseer. Syself, in haar trots, het gedink dat sy
baie sterk was; nou voel sy dat, as sy die respek van haarself en dié
van Jan wil verdien, sy sonder swakkerigheid die bitterheid en
opstandigheid uit haar hart moet ruk wat vir hulle daar opgehoop het
soos dooie blare in ’n bos in die winter.
—Gaan jy nou al weg? vra mevrou Kibert, half verdrietig.
Jan wil haar troos, deur met haar te praat oor al die bande wat
haar nog aan die lewe bind. Hy doen navraag na haar ander kinders:
Magriet, wat sendelinge is, en haar seuns in Asië.
—Hoeveel kinders het Etienne al?
—Die derde word verwag. Ek ken nog geeneen van hulle nie,
maar o, ek het hul lief, as die laaste blydskap wat God my nog skink.
Hul name is Mouritz en Francina, na my man en na my genoem. Dié
wat kom, sal Marcel genoem word.
—En as dit ’n dogtertjie is?
—Dan Marcella. Kyk, hier is die portret van die twee oudste.
—Is hulle nie fraai nie? sê Paula.
—Ja, die meisie lyk na jou, sy het jou swart oë.
—O, sy sal baie mooier wees.
—Dit glo ek nie, sê die jonkman, terwyl hy die portret aan mevrou
Kibert teruggee; en met die mooi glimlag, wat sy gesig so jonk laat
lyk, sê hy nog daarby:
—Is dié dan nie mooi genoeg nie? Jy is moeilik tevrede te stel.
Onwillekeurig kry Paula ’n kleur, nuwe lewe vlam op in haar bleek
gelaat en gee haar ’n ander aansien—so maak ’n sonstraal met ’n
diamant. In haar wanhoop het sy reeds alle besef van haar
skoonheid uit die oog verloor, en met vreugde vind sy dat sy dit nou
weer besit.
Jan is bly om te sien dat die twee hul smart ’n oomblik vergeet, so
hou hy aan met vra omtrent die seuns, en waar hul eintlik woon.
—Ag, sê mevrou Kibert, ek kan al die vreemde name nie onthou
nie; Paula kan dit beter doen as ek.
En Paula vertel hoe haar broers daar aangaan. Dit gaan goed,
maar hul het te baie werk, en niemand uit die ou vaderland wil hulle
kom help nie.
Eenvoudig en duidelik het sy alles vertel. Jan is opgetoë:
—Hier is geen toekoms meer vir ’n man nie, ek gaan hulle help.
—En jou militêre loopbaan? sê mevrou Kibert, terwyl die jonkman
opstaan om weg te gaan.
—Ek het daar nie so ’n lus in soos Marcel nie; ’n mens verlies so
baie tyd en krag. Maar nou klets ek, mevrou.
Hulle stap uit op die stoep; die front van die huis is toe onder die
blare en blomme van kamperfoelie, rose en meelkruid. Hulle leun
teen die stoeptralies. Hierdie lentemôre is ’n fees vir alle oë wat op
haar skoonheid rus. Die lug is deurskynend en yl. ’n Blouagtige waas
—teken van mooi weer—verdowwe sag die grillige vorme van die
berge. Onder in die laagte skuil die kerktorinkie van die dorpie weg
in die skaduwee van die kastaiingbome. Naderby het die veld die
fris-heldere groen kleur wat net in die voorjaar gesien word. Die
halme van die jong gesaaides tril in die windjie wat verbydwaal; in

Textbook Group Processes Data Driven Computational Approaches 1St Edition Andrew Pilny Ebook All Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Textbook Group Processes Data Driven Computational Approaches 1St Edition Andrew Pilny Ebook All Chapter PDF

Uploaded by

Copyright:

Available Formats

Group Processes Data Driven

Computational Approaches 1st Edition

Monitoring Multimode Continuous Processes: A Data-

Computational Learning Approaches to Data Analytics in

Healthcare Big Data Analytics Computational

Data Driven Approaches for Healthcare-Machine learning

Data Driven Treatment Response Assessment and Preterm

Data Driven Fault Detection for Industrial Processes

Statistical Process Monitoring Using Advanced Data-

Group processes : dynamics within and between groups

More information about this series at http://www.springer.com/series/11784

ISSN 2509-9574 ISSN 2509-9582 (electronic)

Library of Congress Control Number: 2017930624

© Springer International Publishing AG 2017

Printed on acid-free paper

This Springer imprint is published by Springer Nature

Andrew Pilny and Marshall Scott Poole

© Springer International Publishing AG 2017 1

Andrew Pilny and Amanda R. Slone

2.1 Introduction to Response Surface Methodology

A. Pilny (*) • A.R. Slone

© Springer International Publishing AG 2017 5

2.2 Brief Background of RSM

2.3 Basic Processes Underlying RSM

2.3.1 Step 1: Second-Order Regression Modeling

In this general linear equation, y equals a response variable, x1 and x2 represent

2.3.2 Step 2: Lack of Fit

Likewise, if the test is statistically significant for a second-order model, then by

2.3.3 Step 3: Coding of Variables

Salt value - (10 - 2 )

2.3.4 Step 4: Canonical Analysis of the Response System

and then solve for the stationary point:

In these equations, b equals a vector of first-order beta coefficients:

é 3.65 ù  é -1.22 -0.25 ù

B = matrix ( c ( 1.22, - .25, - .25, - 2.66 ) , nrow = 2, ncol = 2, byrow = TRUE )

y = 69.65 + 12.22 x1 + 3.45 x2 - 9.33 x12 - 6.32 x22 - 4.65 x1 x2

y = ys + l1 w12 + l2 w22

= 63.63 + 1.34 w12 + 0.75w22

By taking the determinant of the matrix:

( -1.22 - l ) ( -2.66 - l ) - ( -0.25 * -0.25) = 0

2.3.5 Step 5: Conduct Ridge Analysis if Needed

y = 69.65 - 14.87 x1 + 7.94 x2 - 0.33 x12 - 8.89 x22 + 13.65 x1 x2

and the resulting response in canonical form was

Table 2.1 Example ridge analysis table

2.4 RSM in Context

To demonstrate the usefulness of RSM in group research, this exemplar study

2.4.1 About the Game

2.5 Dependent Variable

2.5.1 Team Performance

2.6 Independent Variables

task complexity given by Wood (1986). According to Wood (1986), complexity

2.7 Control Variables

2.7.1 Group Size

2.8 Data Analysis

2.8.1 Controlling for Group Size

2.8.2 Experience Points: A Minimum Stationary Point

The next line of code runs the RSM procedure:

The SAS System

The RSREG Procedure

Coding Coefficients for the

Factor Subtracted off Divided by

Response Surface for Variable experience_pts:

Regression DF Type I Sum of Squares R-Square F Value Pr > F

2.1 Introduction to Response Surface Methodology

2.2 Brief Background of RSM

2.3 Basic Processes Underlying RSM

2.3.1 Step 1: Second-Order Regression Modeling

2.3.2 Step 2: Lack of Fit

2.3.3 Step 3: Coding of Variables

2.3.4 Step 4: Canonical Analysis of the Response System

é 3.65 ù é -1.22 -0.25 ù

y = 69.65 + 12.22 x1 + 3.45 x2 - 9.33 x12 - 6.32 x22 - 4.65 x1 x2

y = ys + l1 w12 + l2 w22

2.3.5 Step 5: Conduct Ridge Analysis if Needed

y = 69.65 - 14.87 x1 + 7.94 x2 - 0.33 x12 - 8.89 x22 + 13.65 x1 x2

2.4 RSM in Context

2.4.1 About the Game

2.5 Dependent Variable

2.5.1 Team Performance

2.6 Independent Variables

2.7 Control Variables

2.7.1 Group Size

2.8 Data Analysis

2.8.1 Controlling for Group Size

2.8.2 Experience Points: A Minimum Stationary Point

2.9.1 Model for Deaths: A Saddle Point