t tests in practice

T- Tests in Practice

The research article explores application of statistical tests of significance tools in establishing cause-effect relationship. In this case, t-test is the tool used by the author of the article to address the topic on Premorbid Intelligence Estimate Method. Critical evaluation of this article gives an insight into the relevance of various statistical tools for making references over observed relationship among variables. In this study, relevance, validity, and reliability of CPIE in predicting Full Scale IQ (FSIQ), was tested. To establish functionality of CPIE, the authors identify a hypothesis. A null hypothesis for this article has been based on two samples. In the case with a hypothesis and a sample size of more than 30, t-test is more reliable hence it was applied by the author.

The multiples of variables that include ethnicity, level of education of parents, age, and region were used to prove the hypothesis was imperative. In fact, there were mainly two null hypotheses. The first one was that estimated premorbid FSIQ scores would show significant difference from the scores that would be attained from traumatic brain injury (TBI) victims. The second is that estimated premorbid FSIQ score would not be different from the scores obtained from healthy children. Since the sample from children who had suffered from TBI is independent from the sample of healthy students, two-sample t-test was the most appropriate tool. Considering the article attempts to confirm if there is any similarity in mean score of the same variable between the two independent population samples, t-test would yield more accurate and reliable data. Besides, the author’s use of the same sample size of 40 for each sample provided a level ground that would help to minimize occurrence of type II error.

It is imperative that data is shown while using any statistical test of significance. In this case, the author provides all necessary data. However, it is evident that methodology is not covered in its entirety. For each sample, the size is 40 children aged from 6 to 16 years. Besides, the author gives the average age of the sample population at 13.3 with a narrow standard deviation of 2.9. It can also be noted that ethnic breakdown as a percentage of each sample is provided among other variables. The author even gives a list of parameters used to determine the sample sizes, level of significance, and alpha at 0.01.

The tabulated analyzed data findings provide supportive facts to the conclusions. In that respect, the article satisfies the characteristics of a quantitative research. Most of the arguments are reinforced by numerically provided, analyzed, and presented data. Besides, there is an elaborate tabular representation of research findings that informs the direction of the hypothesis testing.

The question as to whether the results stand alone is taken care of by the exhaustive presentation of analysis in table form. Although analysis process that may have included use of SPSS is not shown, the result is sufficient to justify provided conclusions.  Besides, the t-test was used to estimate p-values that are invaluable in this research. Clearly explained research design, parameters, and all variables are a proof that the results are fully supported. Besides, theoretical discussions that derive most of its arguments from literature review provide support for the methodology, results, and conclusions. The results are therefore fully supported; factual evidence from several other scholarly findings is referenced in the article.

 

 

 

Get a 10 % discount on an order above $ 100
Use the following coupon code :
SKYSAVE