Page 65 - Quantitative Data Analysis

Quantitative Data Analysis
Simply Explained Using SPSS

If sample size of the two groups is not equal (n 1 ≠ n 2) then

̅ − ̅ √ ×

Thus, the general formula to calculate the t-test statistics or t

observed value is

̅ ̅

− −

√ ×
−

Assumptions for Independent Sample t-Test

1. Samples must be independent

If samples are not independent but grouped in some way, cannot
separate the effect of the treatment from the effect of the groups.

Example: To compare two classes of 30 students where one group
uses the blue textbook and the other uses the red textbook, it is not
possible to separate the effect of the book from the effect of the
teacher. In such cases, use classrooms as the unit of analysis.

2. Data are interval or ratio level

Without this level of data, computing means and standard
deviations is not meaningful. Nonparametric tests would be an
alternative for nominal or ordinal data.

The Theory and Applications of Statistical Inferences 49