How to Determine the Independent Sample T-Test using SPSS Platform

Research of Independent Sample T-Test and Writing Anxiety Result for XI Class

An Independent Sample t-test is a statistical analysis technique useful for seeing the relationship between variables.  The difference test is used to determine significant differences between conditions before and after a treatment and there are differences between two samples.  The difference test can also be used to see the effect of a treatment.

In a research case, we are often faced with a relationship between an independent variable that is categorical or on a non-metric scale and a dependent variable that is continuous or on an interval/ratio scale.  The analysis technique that is suitable for this case depends on the number of independent variable categories.

Introduction

If the independent variable has 2 categories, then the statistical test that is suitable to use is the t-test.  Meanwhile, if there are more than 2 categories, the statistical test that is suitable to use is an analysis of variance (ANOVA).  However, if there is more than one dependent variable then what is suitable to use is a multivariate analysis of variance (Manova) (Ghozali, 2013: 63)

according to (Feng et al., 2017) the t-test for independent samples is a type of parametric inferential statistics (differential test or comparison test).  The one-sample t-test is mainly used to compare the sample mean with the mean of a particular population.

According to Santoso (2008:211), basically, the two-sample test is to find out, if there is a difference in the average (mean) between two populations, by looking at the average of the two samples.

 

Case :

A researcher conducted research on "Student Anxiety in Writing".  This research was conducted to test whether there was a significant difference between the writing anxiety of students in class A and class B. The following table lists the recapitulation of the writing anxiety test results for class A and class B.

Daftar Rekapitulasi Hasil Tes Kecemasan Menulis Kelas A dan Kelas B

No.	Skor Kelas A	No.	Skor Kelas B	Kode Kelas A	Kode kelas B
1	76	1	71	1	2
2	84	2	76	1	2
3	94	3	64	1	2
4	84	4	71	1	2
5	94	5	64	1	2
6	87	6	64	1	2
7	94	7	71	1	2
8	84	8	54	1	2
9	94	9	76	1	2
10	84	10	84	1	2
11	94	11	76	1	2
12	96	12	84	1	2
13	94	13	76	1	2
14	87	14	64	1	2
15	64	15	76	1	2
16	84	16	94	1	2
17	87	17	64	1	2
18	87	18	71	1	2

 

 

 

 

 

 

 

 

 Steps for Independent Sample t-test (SPSS v.20)

• on the Home page of SPSS, Click File-Open-Data to activate the file
• choose/click the file then Ok 

• To make the testing process easier, please enter the SPSS variable view enter the value 1 in values ​​and class A in Labels then add similarly for class B enter the value 2 in values ​​and CLASS B in Labels

• on the main menu of SPSS (Data View) Click Analyze - Compare Means - Independent Sample T Test
• Fill in the result variable of writing anxiety in this example
• Fill the class to Grouping Variable column
• Click Define Groups, if not activated, you need to click Define Groups first until the color turns to yellow, then click Define Groups again
• add number 1 in Group 1 then add number 2 in Group 2 to mark that there are 2 independent samples, like class A and B, then click OK, and the result looks like this:

Output Analyzing:

First Output analysis in statistics group

In the writing anxiety test results and the mean column, you can see that the average writing anxiety test result for class A is 85.7241, while the average result for the writing anxiety test for class B is 70.5517.

 It can be seen that from the characteristics of the writing anxiety test results class A and class B have different averages.

 To see whether the difference is statistically real (significant), we need to look at the output of the second part (Independent Sample t-test)

Second Output Analysis on Independent Sample T-Test

Two stages of analysis must be carried out in this second part, namely:

 1. First test the assumption whether the population variance of the two samples (men and women) is the same (equal variance assumed) or different (equal variances not assumed)

 2. After knowing whether the variance is the same or not, then look at the t-test value to determine whether there is a real (significant) difference in the average value or not

 The test hypothesis for each variable is as follows:

 HO: The population variance of the writing anxiety test results between class A and class B is the same

 H1: Population variance in writing anxiety test results between class A and class B is different.

 

Decision-making:

 If the probability is > 0.05, then H0 cannot be rejected so the variance is the same

If the probability <0.05, then H0 is rejected so the variance is different

Analysis Results

 It can be seen in the second table (independent samples test) that the anxiety test result variable, writing the F value of the Levene test of 0.015, has a probability (sig.) greater than 0.05 (0.903>0.05), so it can be concluded that H0 cannot be rejected.  Thus, the analysis of different tests (t-test) must use the equal variance assumption.  The t value for equal variance assumed is 6.368 with a significance probability of <.001(<.001<0.05)(two tail).  So it can be concluded that the average results of the class A writing anxiety test and the average results of the class B writing anxiety test are different (significantly different).

 That's how to do a difference test or independent sample t-test.  Easy isn't it?!  For those who want to share comments, please send them in the comments column.  For those who want to subscribe, please send an email.  Thank you for your visit, I hope that what the author explains above is useful for all my friends.  Wassalam.