ANOVA TEST to Determine Statistical Significance

Running ANOVA test to check whether the total value of ATM transactions using debit cards in India is influenced by the month of the year. The data for the analysis has been obtained from the RBI (Reserve Bank of India) database and can be accessed here:

https://www.rbi.org.in/Scripts/ATMView.aspx

In order to make sure the data is in a format fit for analysis, I have reformated the data to look like the following:

http://falcon80.com/Blogs/wp-content/uploads/2016/03/ATMTRANSACTIONS.xlsx

Null Hypothesis: The value of ATM transactions is not influenced by the time of the year.

SAS Code for the ANOVA Test:

/* Generated Code (IMPORT) */
/* Source File: ATMTransactionsIndia.xlsx */
/* Source Path: /home/isabelleraja0 */
/* Code generated on: Sunday, March 13, 2016 2:13:24 AM */

/* Generated Code (IMPORT) */
/* Source File: ATMTRANSVALUE.xlsx */
/* Source Path: /home/isabelleraja0/sasuser.v94 */
/* Code generated on: Sunday, March 13, 2016 10:50:31 PM */

%web_drop_table(WORK.IMPORT);

FILENAME REFFILE “/home/isabelleraja0/sasuser.v94/ATMTRANSACTIONS.xlsx” TERMSTR=CR;

PROC IMPORT DATAFILE=REFFILE
DBMS=XLSX
OUT=WORK.IMPORT;
GETNAMES=YES;
PROC CONTENTS DATA=WORK.IMPORT;

%web_open_table(WORK.IMPORT);
RUN;
LABEL AVERAGE= “Average Value of ATM debit Card Transactions”;

PROC ANOVA; CLASS Month;
MODEL AVERAGE = Month;
MEANS Month;

RUN;

Output:

Dependent Variable: AVERAGE AVERAGE

Source DF Sum of Squares Mean Square F Value Pr > F
Model 11 141511021368 12864638306 0.10 0.9999
Error 44 5.4603717E12 124099357802
Corrected Total 55 5.6018828E12
R-Square Coeff Var Root MSE AVERAGE Mean
0.025261 22.24858 352277.4 1583370
Source DF Anova SS Mean Square F Value Pr > F
Month 11 141511021368 12864638306 0.10 0.9999
Distribution of AVERAGE by Month

The ANOVA Procedure

Distribution of AVERAGE by Month
Level of
Month
N AVERAGE
Mean Std Dev
April 5 1536279.49 375268.096
August 5 1584656.82 379527.519
December 4 1575451.39 284465.328
February 4 1461088.70 203371.834
January 4 1587887.78 307202.647
July 5 1599382.48 379468.261
June 5 1550018.21 354715.995
March 4 1664288.02 290951.425
May 5 1606618.10 391837.800
November 5 1597240.90 366743.451
October 5 1663292.20 388804.856
September 5 1565286.56 381076.242

Inference:

Since the P value =0.99 and is very high, there is not enough evidence to reject the null hypothesis. Since there is not evidence to reject the null hypothesis, there is no need for a post-hoc test.

This entry was posted in Big Data, Data Analysis, Uncategorized and tagged . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *