Dependent T-Test Analysis

Dependent T-Test Analysis

Introduction

School administrators use different types of disciplinary actions for students to contain deviant behaviours. These actions include office referral, suspension in and out of school, and some form of corporal punishment to promote discipline in students.  This analysis aims to assess the effectiveness of disciplinary action on student discipline. Some pertinent questions explored include: How effective is the use of current disciplinary actions such as office referral and student suspension in fostering student discipline? What is the effect of the intervention on student suspension? The null hypothesis is: there is no significant difference in the suspension rate before and RTI. In particular, the analysis examined the rate suspension before and after RTI (response to intervention) to ascertain the effectiveness of current disciplinary interventions. The dependent t-test is conducted to assess the effectiveness of disciplinary interventions such as student suspension.

Method

             The study targeted male student in grade four from different parts of the district. Secondary data on students’ information such as RTI, suspension rates, grade, and gender from school registers were used in the study. The selection of school included in the study was done using stratified random sampling methods to ensure inclusion of all regions in the sample. Further stratification at school level was based on student race/ethnicity to reduce bias by ensuring that all races/ethnicity are represented in the sample (Alvis, 2016). The variables of interest in this analysis include annual suspension as the dependent variable and RTI as an intervention variable. These variables have continuous data that meet the normality assumption. The collected data were checked for completeness and coded for input into the statistical program for analysis. IBM statistics version 23 was used to analyze collected data based on dependent T-test/paired sample t-test analysis.

Results

Descriptive Statistics

Table 1: descriptive results

As shown in Table 1 above, the descriptive analysis N = 30 male students in grade four with suspension mean of 2.37 before the intervention and 1.3 after an intervention. Descriptive statistics have confirmed existing a difference in mean of suspension rate before and after the intervention. The standard deviation for before and after the response was 3.023 and 2.103, respectively. Furthermore, the standard error was also higher in the pre-intervention group than the post-intervention group at .552 and .384, respectively. Accordingly, the post-intervention group had a lower mean, standard deviation, and standard error mean value than the control group or pre-intervention group. It is essential to determine whether a difference in mean value between the two groups is statistically significant using a paired sample test/dependent t-test.

Dependent T-test

            The paired T-test reveals a mean of 1.06, which is the mean difference between post-intervention and pre-intervention groups. The standard deviation and standard error are 1.337 and .244, which is the difference between the pre and post-suspension groups. The t-value is 4.369, much high than the t alpha for N= 30. Most importantly, the P-value is .000, which is less than 5 per cent. Therefore, we reject the hypothesis that no significant difference exists between suspension before RTI and suspension after RTI. In conclusion, the difference in mean suspension rates between the two groups is substantial. Therefore, current disciplinary intervention is effective in promoting discipline among male grade students as evidenced by the reduction of mean suspension rate between pre- and post-intervention groups.

Table 2: Paired Sample T-test results

Calculation and interpretation of Effect size

The effect size is computed calculating using Cohen’s deviation denoted as d. A large d value indicates a strong effect size, while weak d values indicate a weak effect size. The value of deviations d= .2 represents a small effect size, d = .5 show a moderate effect size while a large effect size has a value of more than 0.8 standard deviations (Field, 2018).

Calculation:

                         d=t √N1+N2/NIN2

= 4.369 √30+30/30*30

=4.369√60/900

=4.369√0.0667

= 4.369*.258

d = 1.127

Interpretation

The effect size is 1.127, meaning that a large standard deviation exists in the pre and post-test groups. The rate of student pre- suspensions is 1.127 standard deviations higher than suspension rate post-RTI. Since the d value is greater than 0.08, differences in means between the pre- RTI and Post RTI has great effect size.

Discussion

The study used paired test to analyze data on the effect of the disciplinary intervention on student suspension. The study objective was to assess the difference in the suspension of the student before RTI and after RTI. The tests sought to test the hypothesis that there are no significant differences in pre-intervention and post-intervention groups. Descriptive analysis revealed differences in mean values between pre-RTI suspension and suspension rate after RTI. The dependent test was applied to confirm that a difference in mean value between these groups was significant. The paired T-tests had a p-value of .000; hence the null hypothesis was rejected. This finding has implication on the education sector regarding student discipline in the middle class. It means that current disciplinary action, particularly student suspension, has the potential to reduce the rates of indiscipline in grade four male students.

However, the small sample size used in the study is a significant limitation. In this regard, an enhanced sample is recommended for improved credibility of the results. The reliant on secondary data is another limitation likely to affect the outcome. The study also relied on quantitative method hence s subject to face its drawbacks.  It is therefore vital to obtain primary data and more so qualitative data to obtain insightful information on the topic from education experts. Overall, the study met its purpose by answering the research question and testing the study hypothesis.

References

Alvi, M. (2016). A manual for selecting sampling techniques in research.

Field, A. (2018). Discovering statistics using IBM SPSS statistics (5th ed.). Thousand
Oaks, CA: Sage.