Contact Us

By telephone
During office hours
(Monday-Friday 08:30-17:00)
+44 (0)1234 400 400

Outside office hours
(Campus Watch)
+44 (0)1582 74 39 89

By email
admission@beds.ac.uk (admissions)
international@beds.ac.uk (international)
sid@beds.ac.uk (student support)
help@beds.ac.uk (registration)

By post
University of Bedfordshire
Park Square
Luton
Bedfordshire
UK, LU1 3JU

t-test

t-test

1. Aims and Learning Objectives

  • The aims of this workshop are to use SPSS to carry out an Independent samples T-test and a Related samples T-test.

After this workshop you will be able to:

  • Use SPSS to carry out an Independent samples T-test and a Related samples T-test.
  • Recognise the difference in the format for data entry between independent and related samples.
  • Produce illustrative statistics for both tests.
  • Copy and paste the illustrative statistics into Microsoft Word.

2. What is a T-test?

The T-test is used to decide whether two samples are statistically different from each other or not. It examines the differences in the sample means and allows us to decide whether we would expect this significance by chance or whether it indicates the samples probably come from different populations.

There are two main things that you must check before carrying out a T-test:

  1. The t-test is parametric which means it can only be performed on interval or ratio data, and that data must be normally distributed. For Nominal or Ordinal data you must carry out a non-parametric test (Wilcoxon or Mann-Whitney U).
  2. Are your samples related or Independent? Although, you can do a T-test on both types of design you must know which design you are using as a different formula for related T-test and the Independent T-test.

3. Related samples T-test

Before conducting the related T-test, you need to check that your data fully fits the requirements for a parametric test, i.e. normally distributed and the data is at least of interval status. For the related T-test you must also ensure that your data is of a repeated measures design, i.e. each participant was his or her own control.

t-test
  • N. B. Note that compared to the Independent Data set the format is different (i.e. there is no independent variable).
T-testT-test

These tables should now appear in the Output window.

T-test
  • The Paired Samples Test gives the inferential statistic.
  • It may not be the easiest table to read but there are 3 main items that you need to take from it namely the t value, df, and sig. value.
T-test
  • You may have noticed that SPSS has automatically chosen a two-tailed significance level. There is not an option to choose a 1 tailed test for the T-test. However, if your hypothesis is 1-tailed, all you have to do is divide the two-tailed value by 2.

e.g. 0.005 becomes 0.0025

Task 1: Remind yourself of the Hypothesis and, write down the values for t, df and the P value. Are you going to accept or reject the null hypothesis?

4. Illustrative Statistics for T-tests

You will recall from your lectures that the illustrative statistics for means are bar charts. As the T-test compares means this is the most appropriate illustrative statistic to use.

Task 2: To produce a bar chart for this data.

T-testT-test

5. Independent/Unpaired T-test

If you need to compare the means of two groups of participants for the same variable then you need look no further than the Independent T-test. E.g. In a class of 14 students, 7 attended regularly whilst only 7 attended 50% of their classes. Therefore we can split the class into two groups.

H1 – Students who attend class regularly will achieve higher exam results than those who do not.

T-testT-test
  • Click on continue, and then OK.
  • These tables will then appear in the Output window.
T-test
  • The next table is the Independent-samples table, shown below.
  • It has a lot of information on it.
  • The main part of the table to concentrate on is the t value, df, & Sig., 2 tailed.
T-test
  • You will notice that in each of these boxes there are two values. One value is for equal variances assumed and the other is for equal variances not assumed. To help you to decide which one to use, look at the Levene’s Test for Equality of Variances. If this test is not significant at p<0.05 then use the figures on the top line. If it does show a significant difference then use the figures on the second line.

Task 3: From the values given, decide whether or not you are going to accept or reject the null hypothesis.

Task 4: Create a bar chart to show the illustrative statistics.

(N.B. This procedure is slightly different when the design is independent measures due to the different format of the data. Select Summaries for groups of cases’.

1) Click on the ‘Other summary function’ radio button and send ‘exam’ to the variable box. Then send ‘group’ to the ‘Category Axis’ box. Now click OK.

T-test

2) Copy and paste tables into word, then save your work.

Bedfordshire University

Apply» Faculties & Departments» Department of Psychology» Labs» SPSS statistical procedures» t-test

Top