You are on page 1of 54

Methodology of Conducting RESEARCH

TECK (PVT)

and Elaborating upon LTD

Interpretations of
Quantitative Research
Techniques using SPSS
Inquiry Mode

 Qualitative Research
 Quantitative Research
Qualitative & Quantitative
Research
“Not everything that can be counted counts and
not everything that counts can be counted.”

( Albert Einstein )
What is qualitative research?
• Qualitative research seeks out the ‘why’ and ‘how’ of its
topic through the analysis of unstructured information –
things like interview transcripts, emails, notes, feedback
forms, photos and videos. It doesn’t just rely on statistics or
numbers, which are the domain of quantitative researchers
• Qualitative research is an approach which seeks to
understand , by means of exploration, human experience,
perceptions, motivations, intentions and behavior
• It is interactive, reflexive, inductive, flexible, and holistic
method of data collection and analysis
(Bryman & Bell 2007)
What is quantitative research?
• Quantitative Research seeks out the ‘how many’, when,
‘where’ of its topic through the analysis of structured
information. It relies on statistics or numbers
• In Quantitative Research we try to quantify the variation in
the phenomenon, situation, problem or issue. Statistical
analysis is done to determine the magnitude of variation
When Should I Use Qualitative Vs. Quantitative Research- - YouTube_2.flv

Qualitative and Quantitative Approaches


Qualitative Quantitative
(Usually) Non-probability based Typically a probability-based
sample sample
Non-generalizable Generalizable

Answers Why? How? Answers How many? When?

Formative, earlier phases Tests hypotheses, latter phases

Data are “rich” and time- Data are more efficient, but may
consuming to analyze miss contextual detail
Design may emerge as study Design decided in advance
unfolds
Researcher IS the instrument Various tools, instruments
employed
The Main Steps in Quantitative Research
1. Theory

2. Hypothesis (deductive stage)

3. Research design

4. Derive measures of concepts

Feedback 5. Select research sites

(inductive 6. Select research subjects/respondents


stage)
7. Administer research instruments/collect data

8. Process data

9. Analyse data

10. Findings/conclusions

11. Write up findings/conclusions Fig. 6.1


Why Measure?
• To delineate fine differences between people,
organizations, or any other unit of analysis

• As a consistent device for gauging distinctions

• To produce precise estimates of the degree of


relationship between concepts
Why Use More Than One Indicator?

• Single indicators may incorrectly classify many


individuals
• Single indicators may capture only a portion of
the underlying concept or be too general
• Multiple indicators can make finer distinctions
between individuals
• Multiple indicators can capture different
dimensions of a concept
 Psychometric Tests

 Reliability
(Cronbacks’ Alpha α)
 Validity
(Factor Loadings, CFA & EFA)
Types of Reliability
• Stability
▫ is the measure stable over time?
 e.g. test–retest method
• Internal reliability
▫ are the indicators consistent?
 e.g. split-half method
• Inter-observer consistency
▫ is the measure consistent between observers?

see Research in focus 6.5 &


6.6
Types of Validity
• Face validity
• Concurrent validity
• Predictive validity
• Construct validity
• Convergent validity
see Key concept 6.7
Face Validity
A researcher developing a new measure could establish
it has face validity - i.e., that the measure reflects
the content of the concept in question. This might
be established by asking other people whether the
measure seems to be getting at the concept that is the
focus of attention. E.g. people, possibly with experience
or expertise in a field, might be asked to act as judges to
determine whether on the face of it the measure seems
to reflect the concept concerned. Face validity is,
therefore, an essentially intuitive process.
Concurrent Validity
Researchers can also gauge the concurrent validity of measures,
employing a criterion relevant to the concept in question on
which cases (e.g. people) are known to differ. For a new measure of
job satisfaction a criterion might be absenteeism, some people
being absent from work (other than through illness) more often
than others. Thus we might see how far people who are satisfied
with their jobs are less likely to be absent than those who are not
satisfied. A lack of correspondence, such as there being no
difference in levels of job satisfaction among frequent absentees,
might cast doubt on whether our measure is really addressing job
satisfaction.
Predictive Validity
Another possible test for the validity of a new measure is
predictive validity, whereby the researcher uses a future
criterion measure, rather than a contemporary one, as in the case
of concurrent validity. With predictive validity, the researcher
would take future levels of absenteeism as the criterion against
which the validity of a new measure of job satisfaction would be
examined. The difference from concurrent validity is that a future
rather than a simultaneous criterion measure is employed.
Construct Validity
Researchers could also estimate the construct validity of a
measure, deducing hypotheses from a theory relevant to the
concept. E.g. drawing on ideas about the impact of technology on
the experience of work a researcher might anticipate that people
satisfied with their jobs are less likely to work on routine jobs;
those not satisfied are more likely to work on routine jobs. We
could investigate this by examining the relationship between job
satisfaction and job routine. However, some caution is required as
either the theory or the deduction that is made from it might be
misguided, or the measure of job routine could be an invalid
measure of that concept.
Convergent Validity
The validity of a measure could be gauged by comparing it to
measures of the same concept developed through other methods.
For example, if we develop a questionnaire measure of how much time
managers spend on various activities (such as attending meetings,
touring their organization, informal discussions, etc.), we might
examine its validity by tracking a number of managers and using a
structured observation schedule to record how much time is spent in
various activities and their frequency.
EFA & CFA
Basic Terminologies
 Concept (are categories for the organization of ideas and
observations)
Motivation to work
Charismatic Leadership
Competitive success
Job Satisfaction
 Variable
 Dependent Variable
 Independent Variable
 Measure & Indicator
 Dimensions & Facets
 Items
 Antecedents & Out Comes
 Construct & Model
Level of Significance
P Value
Normality
Scales of Measurement
Dichotomous
Nominal
Ordinal
Interval/Ratio
Determine the Relevant Variable
Anything that may assume different numerical values
• Dependent
• Independent
• Moderating
• Intervening
Independent variable Dependant Variable

Salary Structure Job Satisfaction


Price

Quality

Customer Satisfaction
Brand

Services

Independent variables Dependent variable


Salary Structure Job Satisfaction

Independent variable Dependant Variable

Culture

Moderating variable
Number of Reading
Books abilities

Independent variable Dependant Variable


Number of Reading
Books abilities
Independent variable Dependant Variable

Parents’
Literacy

Moderating variable
Workforce
diversity Organizational
effectiveness

Independent Dependent
variable variable
Workforce Creative Organisational
diversity synergy effectiveness

Independent Intervening Dependent


variable variable variable
Workforce Creative Organisational
diversity synergy effectiveness

Independent Dependent
variable Intervening
variable variable

Managerial
expertise

Moderating Variable
Differentiating between Different types of
Quantitative Analysis

1. Content Analysis
2. Conducting Secondary Data Analysis and
Cross Cultural Comparisons
3. What is Meta Analysis
4. Primary Data Analysis
Types of Quantitative Analysis

1. Uni-variate Analysis
2. Bivariate Analysis
3. Multivariate Analysis
Univariate Analysis Authored by
David McHugh

Univariate analysis refers to the analysis of one variable at a


time. The commonest approaches are as follows:

• Frequency tables • Measures of


central tendency:
• Diagrams: ▫ Arithmetic mean
▫ Bar charts ▫ Median
▫ Pie charts ▫ Mode
▫ Histograms
• Measures of
dispersion:
▫ Range
▫ Standard deviation
Frequency Table Showing Reasons for Visiting
the Gym
Reason n per cent

Relaxation 9 10

Maintain or improve fitness 31 34

Lose weight 33 37

Build strength 17 19

TOTAL 90 100

Table 14.2
Frequency Table Showing Ages of Gym
Members
Age n per cent

20 and under 3 3

21–30 39 44

31–40 23 26

41–50 21 24

51 and over 3 3

TOTAL 89 100

Table 14.3
Bar Chart Showing Main Reasons for Visiting
the Gym
35
30
25
Count

20
15
10
5
0
Relaxation Maintain or Lose weight Build strength
improve fitness
Main reasons for visiting the gym

Fig. 14.2
Pie Chart Showing Main Reasons
for Visiting the Gym

Build strength Relaxation


19% 10%

Maintain or
improve fitness
34%
Lose weight
37%

Fig. 14.3
Histogram Showing the Ages of Gym Visitors

45
40
35
30
Count

25
20
15
10
5
0
20 and 21--30 31--40 41--50 51 and
under over
Ages of Gym Visitors

Fig. 14.4
Bivariate Analysis
• Bivariate analysis is concerned with the
analysis of two variables at a time in order to
uncover whether the two variables are related

• Main types:

▫ Contingency tables
▫ Pearson’s r
▫ Spearman’s rho
▫ Phi and Cramér’s V
▫ Comparing means and eta
Methods of Bivariate Analysis
Nominal Ordinal Interval/ratio Dichotomous
Nominal Contingency table Contingency table Contingency table Contingency table
+ chi-square (χ2) + chi-square (χ2) + chi-square (χ2) + chi-square (χ2)
+ Cramér’s V + Cramér’s V + Cramér’s V. + Cramér’s V
If the interval/ratio
variable can be
identified as the
dependent variable,
compare means + eta

Ordinal Contingency table Spearman’s rho (ρ) Spearman’s rho (ρ) Spearman’s rho (ρ)
+ chi-square (χ2)
+ Cramér’s V

Interval/ratio Contingency table Spearman’s rho (ρ) Pearson’s r Spearman’s rho (ρ)
+ chi-square (χ2)
+ Cramér’s V.
If the interval/ratio
variable can be
identified as the
dependent variable,
compare means +
seta

Dichotomous Contingency table Spearman’s rho (ρ) Spearman’s rho (ρ) phi (φ)
+ chi-square (χ2)
+ Cramér’s V
Fig. 11.5
Contingency Table Showing the Relationship
Between Gender and Reasons for Visiting the Gym
Reasons Gender

Male Female

No. % No. %

Relaxation 3 7 6 13

Fitness 15 36 16 33

Lose weight 8 19 25 52

Build strength 16 38 1 2

TOTAL 42 48

Table 14.4
Comparing Subgroup Means: Time Spent on
Cardiovascular Equipment by Reasons for Going to the
Gym

Time Reasons
Relaxation Fitness Lose Build Total
weight strength
Mean number of 18.33 30.55 28.36 19.65 26.47
minutes
spent on
cardiovascular
equipment
n 9 31 33 17 90

Table 14.5
Three Contexts for Multivariate Analysis

• Mutivariate analysis entails the simultaneous analysis of


three or more variables. There are three main contexts
within which multivariate analysis might be employed:

1. Could the relationship be spurious?

2. Could there be an intervening variable?

3. Could a third variable moderate the relationship?


A Spurious Relationship

Leadership style

Employee Job
commitment satisfaction

Fig. 14.10
Intervening Variables

• Piercy, Haris, and Lane (2002) suggest that higher


levels of market orientation are associated with
higher levels of employee motivation, satisfaction,
and commitment, which in turn leads to enhanced
organizational performance. Employee attitudes are
thus an intervening variable:

market orientation → employee attitudes → organizational performance


Statistical Significance
• The chi-square test
▫ establishes how confident we can be that there is a
relationship between the two variables in the population

• Correlation and statistical significance


▫ provides information about the likelihood that the coefficient
will be found in the population from which the sample was
taken

• Comparing means and statistical significance – the F


statistic
▫ expresses the amount of explained variance in relation to the
amount of error variance
Steps in Statistical Significance
Testing
1. Set up a null hypothesis

2. Decide on an acceptable level of statistical


significance

3. Determine whether findings reach the acceptable


level

4. If acceptable level attained, reject null hypothesis.


If not attained, accept null hypothesis
What Is the Level of Statistical
Significance?
The level of statistical significance is the level of risk
that you are prepared to take that you are inferring
there is a relationship between two variables in the
population from which the sample was taken when in
fact no such relationship exists. The maximum level of
risk that is conventionally taken in business and
managerial research is to say that there are up to 5
chances in 100 that we might be falsely concluding that
there is a relationship when there is not one. This means
that, if we drew 100 samples, we are recognizing that as
many as 5 of them might exhibit a relationship when
there is not one in the population. Our sample might be
one of those five, but the risk is fairly small.

see Box 14.6 for more


Type 1 and Type 2 Errors
Error

Type I Type II
(risk of rejecting the (risk of confirming the
null hypothesis when null hypothesis when
it should be confirmed) it should be rejected)

0.05
Greater Lower
risk risk
p level

0.01 Lower Greater


risk risk
Statistical Analysis

Statistical analysis
WHY SPSS?
Q&A

You might also like