Scribd
Upload
12 views17 pages

Independent Sample T Test

The document provides an overview of the independent sample t-test, a parametric test used to compare the means of two groups to determine if they are significantly different. It outlines the assumptions required for the test, the process for conducting it in SPSS, and how to interpret the results, including the use of confidence intervals. Additionally, it mentions the Mann-Whitney U test as an alternative for non-normally distributed data.

Uploaded by

mali1102013
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
12 views17 pages

Independent Sample T Test

The document provides an overview of the independent sample t-test, a parametric test used to compare the means of two groups to determine if they are significantly different. It outlines the assumptions required for the test, the process for conducting it in SPSS, and how to interpret the results, including the use of confidence intervals. Additionally, it mentions the Mann-Whitney U test as an alternative for non-normally distributed data.

Uploaded by

mali1102013
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

Independent sample T test

Definitions
• A two-sample t-test is a parametric test used to estimate whether the mean value
of a normally distributed outcome variable is significantly different between two
groups of participants.
• Used to assess whether two mean values are similar enough to have come from
the same population or whether their difference is large enough for the two groups
to have come from different populations.
• Can be used to analyze data from any type of experimental or non-experimental
study design where the explanatory variable can be classified into two groups, for
example (e.g. drug treatment group vs. placebo group), they may be naturally
occurring (e.g. male vs. female), or they might be beyond the control of the
experimenter (e.g. depressed people vs. healthy people).
• scores are compared between the two groups (the independent variable).
Definitions
• An approximate formula for calculating a t value , when variances are equal is:

With df = n1 + n2 - 2
Definitions
• An approximate formula for calculating a t value , when variances are not
equal is:
Assumptions
• The assumptions that must be satisfied to conduct a two-sample t-test are:
1. The groups must be independent, that is, each participant must be in one group only.
2. The measurements must be independent, that is, a participant’s measurement can
be included in their group once only.
3. The outcome variable must be on a continuous (interval or ratio) scale.
4. The outcome variable must be normally distributed in each group.
5. The variances between groups are approximately equal, that is, homogeneity of
variances (if data fail this assumption an adjustment to the t value is made).
One- and two-sided tests of significance
• When a hypothesis is tested, it is possible to conduct a one-sided (one-tailed) or a
• two-sided (two-tailed) test.
• A one-sided test is used to test an alternative hypothesis of an effect in only one
direction (i.e. mean1 > mean2 or mean1 < mean2).
• A two-sided test is used to test an alternative hypothesis of whether one mean value is
smaller or larger than another mean value (i.e. mean1 ≠ mean2). That is, there is a
difference in either direction between the two populations from which the samples were
selected.
• When a one-sided test is used, the 5% rejection region is placed only in one tail of the
distribution.
• for a two-sided test, 2.5% of the rejection region is placed in the positive tail of the
distribution (i.e. mean1 > mean2) and 2.5% is placed in the negative tail (i.e. mean1 <
mean2).
One- and two-sided tests of significance
Leven’s test
To check wither or not
we have an equal
variances

We test the null


Hypothesis of equal
variances against the
alternative of unequal
variances.

Sig < 0.05 reject , Sig > 0.05 do not


we choose assume reject , we choose
not equal assume equal
variances variances
Confidence intervals
• Confidence intervals are invaluable statistics for estimating the precision
around a summary statistic such as a mean value and for estimating the
magnitude of the difference between two groups.
• For mean values, the 95% confidence interval lower and upper bounds is
calculated as follows:
Confidence intervals
• If a CI contains zero (-ve,+ve). A zero value for a mean difference would
indicate no difference between groups (Do not reject ).
• If a CI doesn’t contains zero (-ve, -ve) or (+ve, +ve), it indicate there is a
difference between groups (reject ).

• In figure 7.8, 0 is no included in (41.536, 207.864) so there is a difference


between groups means.
• We are 95% confident that the mean difference between the two groups lies
between 41.437 and 207.864.
Note:
• Two-sample t-tests are classically used when the outcome is a continuous
variable and when the explanatory variable is binary.

Dependent variable
(Quantitative)

Variables included in test


Independent variable
(grouping variable) 2
categories.
Running independent t-test in SPSS
1. Before doing the test you must check if the
dependent variable is normally distributed as
we learned in chapter 2.
2. Don’t forget to define the values of your
categorical (independent) variable in the
Variable View section.
3. To run an independent sample t test, Go to
Analyze -> Compare Means ->
Independent-Samples T Test (as shown in
figure 7.4).
4. In new window (see figure 7.5), transfer the
dependent variable to Test Variable(s) box
and transfer the independent variable to
Grouping Variable, then click Define
Groups.
Running independent t-test in SPSS
5. In new window (see figure 7.6), enter
your two categories values inside Group
1 and Group 2, then click Continue.
6. In the original window click Ok.
7. Go to the Output window to check your
analysis.
Interpretation of output
• Testing the hypothesis:
: There is no significance difference between the mean
of Anxious & Not Anxious for Reaction time.

• In figure 7.7, we notice that there is a


difference between the means of Anxious Not
Anxious groups. The mean difference value is
124.700 (In figure 7.8).
• In figure 7.8, the p-value for Leven’s test is
0.515 > 0.05, this implies that we choose
assume equal variances (1st row).
• The p-value (Sig) of our T-test is 0.006 < 0.05,
then we reject and conclude statistically that
there is a significance difference between the
mean of Anxious & Not Anxious for Reaction
time.
Presenting data graphically
• You might also want to present
those data graphically. In this case,
we will see how to draw a bar chart
(with error bars to show 95%
confidence intervals):
1. Go to Graph -> Chart Builder as
shown in the figure 7.9.
2. In new window, select Bar (from
list under Choose from) and drag
Simple Bar graphic into empty
chart preview area (as shown in
figure 7.10).
Presenting data graphically
3. Drag your dependent variable
into Y-Axis box (to the left of the
graph) and drag your
independent/categorical variable
to the X-Axis (bellow the graph)
as shown in figure 7.12.
4. Click Ok.
5. From the graph we will notice the
difference between the two
groups clearly.
Mann–Whitney U test
• If our dependent variable is not normally distributed and we want to compare
the two groups we must use Mann-Whitney U test.
• The Mann–Whitney U test is based on ranking the data values as if they were
from a single sample (the data are sorted in order of magnitude and ranked).
• Once the ranks are assigned, they are then summed for each of the groups
and tested.

You might also like