You are on page 1of 4

Mobile payment

Case Study 03 MSMR


/* --- DATA SECTION -----------------------------------------------*/
/* Importing the data set from an excel spreadsheet : adapt the directory path */
PROC IMPORT DATAFILE="/folders/myfolders/Mobile_payment_case_data.xlsx " OUT=Db_m_payment
DBMS=XLSX REPLACE;
SHEET="db_m_payment_short"; * name of the excel spreadsheet;
GETNAMES=YES; * variable names in the first line of the excel spreadsheet;
RUN;

/* Sorting the dataset */


PROC SORT DATA=Db_m_payment;
BY resp_id; * in ascending order of respondent identificator;
RUN;

/* Transforming the dataset */


DATA Db_m_payment; * creates a new dataset named Db_m_payment;
SET Db_m_payment; * puts the dataset Db_m_payment into the new one with the same name (i.e., it
simply modifies the dataset);

* Assigning long labels to the variables;


LABEL UI1 = "I intend to use my smartphone at this shopping stage";
LABEL UI2i = "I plan to use my smartphone at this shopping stage (*** inv. ***)";
LABEL UI3 = "I will use my smartphone the next time I am at this shopping stage";
LABEL Conv1 = "Using my smartphone would allow me to save time during my shopping";
LABEL Conv2i = "Using my smartphone would make my shopping more time consuming (*** inv. ***)";
LABEL Conv3 = "Using my smartphone would be a convinient way to do shopping";
LABEL Pleasure1 = "Using my smartphone for this shopping stage would be enjoyable";
LABEL Pleasure2 = "Using my smartphone for this shopping stage would be entertaining";
LABEL Pleasure3 = "Using my smartphone for this shopping stage would be fun";
LABEL Eco1 = "Using my smartphone would allow me to do my shopping at a lower financial cost";
LABEL Eco2 = "Using my smartphone would allow me to save money";
LABEL Eco3 = "Using my smartphone would allow me to take advantage of promotional offers";
LABEL RiskPrivacy1 = "Using my smartphone would make me lose control over my privacy";
LABEL RiskPrivacy2 = "Using my smartphone would lead to a loss of privacy because my personal data
could be used without my permission";
LABEL RiskPrivacy3 = "Using my smartphone would expose me to the risk of internet hackers taking
control of my personal data";
LABEL RiskFin1 = "Using my smartphone would lead to potential fraud of my bank account";
LABEL RiskFin2 = "Using my smartphone would subject my bank account to financial risks";
LABEL Sexp1 = "I have a great deal of experience with using a smartphone for doing shopping";
LABEL Sexp2 = "I have used very frequently in the past a smartphone for doing shopping";
LABEL Sexp3i = "I am not familiar with the different functionalities of smartphones for doing shopping
(*** inv. ***)";
LABEL Sexp4 = "I frequently update my knowledge about the functionalities of smartphones for doing
shopping";
LABEL Sexp5i = "I do not feel very confident in using a smartphone for doing shopping(*** inv. ***)";

* Reversing responses of inverted items (item "i") -> same direction;


Mobile payment
Case Study 03 MSMR
UI2 = 8-UI2i;
LABEL UI2 = "I plan to use my smartphone at this shopping stage";
Conv2 = 8-Conv2i;
LABEL Conv2 = "Using my smartphone would make my shopping more time consuming";
Sexp3 = 8-Sexp3i;
LABEL Sexp3 ="I am not familiar for the different functionalities of smartphones for doing shopping";
Sexp5 = 8-Sexp5i;
LABEL Sexp5 ="I do not feel very confident in using a smartphone for doing shopping";
RUN;

/* --- ANALYSES SECTION -----------------------------------------------*/


/* Computing basic descriptive statistics */

PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR Ui1-Ui3;
RUN;
PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR Conv1-Conv3;
RUN;
PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR Pleasure1-Pleasure3;
RUN;
PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR Eco1-Eco3;
RUN;
PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR RiskPriv1-RiskPriv3;
RUN;
PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR RiskFin1-RiskFin2;
RUN;
PROC SUMMARY DATA=Db_m_payment PRINT; * computes means, std, min and max and prints the
results;
VAR Sexp1-Sexp5;
RUN;

/* Running Exploratory Factor Analyses */


PROC FACTOR DATA=Db_m_payment METHOD=PRINCIPAL PRIORS=SMC SIMPLE CORR MSA
EIGENVECTORS REORDER ROTATE=VARIMAX;
VAR Ui1-Ui3 Conv1-Conv3 Pleasure1-Pleasure3 Eco1-Eco3 RiskPriv1-RiskPriv3 RiskFin1-RiskFin2 Sexp1-
Sexp5;
Mobile payment
Case Study 03 MSMR
RUN;

/* Running scale measurement reliability test - Here ALPHA option means computing Cronbach's alpha
coefficients (an aggregate measure of the correlation between all items) Alphas should be greater than
0.7; it is an internal consistency test */
PROC CORR DATA=Db_m_payment ALPHA;
VAR Ui1-Ui3;
RUN;
PROC CORR DATA=Db_m_payment ALPHA;
VAR Conv1-Conv3;
RUN;
PROC CORR DATA=Db_m_payment ALPHA;
VAR Pleasure1-Pleasure3;
RUN;
PROC CORR DATA=Db_m_payment ALPHA;
VAR Eco1-Eco3;
RUN;
PROC CORR DATA=Db_m_payment ALPHA;
VAR RiskPriv1-RiskPriv3;
RUN;
PROC CORR DATA=Db_m_payment ALPHA;
VAR RiskFin1-RiskFin2;
RUN;
PROC CORR DATA=Db_m_payment ALPHA;
VAR Sexp1-Sexp5;
RUN;

/* Computing constructs measures. If measures are unidimensional and reliable, we can compute the
aggregate measure of the construct from the item scores. The simplest way is to average the scores
across the items of the same scale*/
DATA Db_m_payment;
SET Db_m_payment;
Ui = mean(Ui1,Ui2,Ui3); * to be coherent, create a Ui variable from the 3 items Ui1 Ui2 and Ui3;
LABEL Ui = "Intention to use in-store m-payment";
Conv= mean (Conv1,Conv2,Conv3); * to be coherent, create a Conv variable from the 3 items Conv1
Conv2 and Conv3;
LABEL Conv= "Perceived Utilitarian benefits -convenience";
Pleasure= mean(Pleasure1,Pleasure2,Pleasure3);* to be coherent, create a Pleasure variable from the 3
items Pleasure1 Pleasure2 and Pleasure3;
LABEL Pleasure = "Perceived hedonic benefits";
Eco= mean(Eco1,Eco2,Eco3);* to be coherent, create a Eco variable from the 3 items Eco1 Eco2 and
Eco3;
LABEL Eco = "Perceived Utilitarian benefits";
RiskPriv= mean(RiskPriv1,RiskPriv2,RiskPriv3);* to be coherent, create a RiskPriv variable from the 3
items RiskPriv1 RiskPriv2 and RiskPriv3;
LABEL RiskPriv = "Privacy risk";
Mobile payment
Case Study 03 MSMR
RiskFin= mean(RiskFin1,RiskFin2);* to be coherent, create a RiskFin variable from the 2 items RiskFin1
RiskFin2 and RiskFin3;
LABEL RiskFin = "Financial risk";
Sexp= mean(Sexp1,Sexp2,Sexp3,Sexp4,Sexp5);* to be coherent, create a Sexp variable from the 5 items
Sexp1 Sexp2 Sexp3 Sexp4 and Sexp5;
LABEL Sexp = "Smartphone based shopping experience";
RUN;

/* Estimating a regression model */


PROC CORR DATA=Db_m_payment; * first, run a correlation analyses to investigate potential
multicolinearity issues, correlations between IV should be small <.4 better;
VAR Ui Conv Pleasure Eco RiskPriv RiskFin Sexp;
RUN;

PROC REG DATA=Db_m_payment;


MODEL Ui = Gender Age /STB; */STB option: standardized coefficient estimates (makes the coefficient
comparable in terms of magnitude);
RUN;
PROC REG DATA=Db_m_payment;
MODEL Conv = Gender Age /STB; */STB option: standardized coefficient estimates (makes the
coefficient comparable in terms of magnitude);
RUN;
PROC REG DATA=Db_m_payment;
MODEL Pleasure = Gender Age /STB; */STB option: standardized coefficient estimates (makes the
coefficient comparable in terms of magnitude);
RUN;
PROC REG DATA=Db_m_payment;
MODEL Eco = Gender Age /STB; */STB option: standardized coefficient estimates (makes the coefficient
comparable in terms of magnitude);
RUN;
PROC REG DATA=Db_m_payment;
MODEL RiskPriv = Gender Age /STB; */STB option: standardized coefficient estimates (makes the
coefficient comparable in terms of magnitude);
RUN;
PROC REG DATA=Db_m_payment;
MODEL RiskFin = Gender Age /STB; */STB option: standardized coefficient estimates (makes the
coefficient comparable in terms of magnitude);
RUN;
PROC REG DATA=Db_m_payment;
MODEL Sexp = Gender Age /STB; */STB option: standardized coefficient estimates (makes the
coefficient comparable in terms of magnitude);
RUN;
QUIT;

You might also like