The Non-Parametric or t-Tests Assessment


Researchers have to use methodical procedures in their studies to guarantee that the targeted spectators recognize the result as convincing. The development of study questions, objectives, sample selection, gathering of facts, and data examination and management enables perfect and suitable conclusion. The t-Test (with t meaning two) is a parametric test used to compare the mean of two groups of data (Plichta, & Kelvin, 2012). Some of the types of t-Test include the independent-samples t-Test, the dependent-sample t-Test, and one sample t-Test (Plichta, & Kelvin, 2012). The use of this parametric test enables researchers to make relevant conclusions in their study.

The use of parametric tests in research is a general finding due to their usefulness in data analysis. Some of the researchers who used the t-Test include Wagner, McDonald, and Castle (2012). They used it in their research article titled, Relationship between Nursing Home Safety Culture and Joint Commission Accreditation, which was published in the fifth issue of volume 38 of The Joint Commission Journal on Quality and Patient Safety. This paper provides an analysis of the study, the use of the t-Test in this study, the statistical assumptions levels of measurement, and the appropriateness of the data display.

t-Test Description

The t-Test family consisting of one sample t-Test, the dependent sample, t-Test, and the independent sample t-Test is an example of parametric tests that are used to compare the means of two groups besides being used at the bivariate level (Plichta, & Kelvin, 2012). When a characteristic is to be measured on a continuous scale, and that two means from the same study are obtained, the independent sample t-Test or the dependent sample t-Test can be used (Plichta, & Kelvin, 2012). The independent sample t-Test is used where the means are from two subgroups of the studied categorical variable (Plichta, & Kelvin, 2012). On the other hand, the dependent sample t-Test is used in comparison with the means of two conditions of the same group such as one characteristic of the group being measured at different times during the study (Plichta, & Kelvin, 2012).

In one sample t-Test, the comparison of means is from two different groups. One of the means is that of the group used in the study while the second mean that is compared to the first mean belongs to another study of the same population (Plichta, & Kelvin, 2012). In the research by Wagner, McDonald, and Castle (2012), the dependent samples t-Test, also known as the paired sample t-Test, is used to compare the means of the two groups of respondents.

Study Description

In the study by Wagner, McDonald, and Castle (2012, p. 207), the researchers were assessing the, “impact of Joint Commission accreditation on patient safety culture perceptions among senior managers in nursing homes in the United States”. The researchers had previously described that there were improvements in safety in hospitals after the intervention in the safety culture in these institutions (Wagner, McDonald, & Castle, 2012). The security background in nursing areas is observed to be inadequately established. This claim is one of the gaps that the researchers wanted to find a way of closing.

The research took place in all the states of the US, with 6,000 nursing homes that were randomly selected participating in the study (Wagner, McDonald, & Castle, 2012). The research involved the sending of a survey to the facilities. The administrators as well as the nursing directors were required to fill the survey (Wagner, McDonald, & Castle, 2012). The results from the institutions that were accredited were compared with those from non-accredited nursing homes, with the paired t-Test being used in the comparison of these differences (Wagner, McDonald, & Castle, 2012).

The researchers later used multivariate analyses to evaluate the association between the safety culture in the nursing homes and the accreditation status of the institutions. They concluded that the accreditation of nursing homes by the Joint Commission was favorable to the safety culture. They recommended accreditation of these institutions to ensure improved safety among the residents (Wagner, McDonald, & Castle, 2012).

Statistic Use in the Study

In the study above, researchers obtained the ratings of the residence safety culture (RSC) in the institutions that were accredited and those that were not certified (Wagner, McDonald, & Castle, 2012). The survey that they had sent to these institutions was analyzed, with the means of the results being calculated. For the study to be relevant, the researchers needed to show that there was a significant difference in the RSC for the two sets of institutions. They utilized the paired sample t-Test to compare ratings from the two groups of respondents.

The researchers obtained positive values, thus indicating that the accredited institutions had relatively higher scores in relation to those that were non-accredited (Wagner, McDonald, & Castle, 2012). They also obtained negative values, which indicated that the non-accredited institutions had higher RSC scores (Wagner, McDonald, & Castle, 2012). Wagner, McDonald, and Castle (2012) needed to determine whether the differences in these values were different from zero at a significant level. Therefore, they used the student’s paired t-Test to determine the significance of the differences.

The results of the t-Test indicated that there were statistically significant differences present between the RSC ratings in the two groups of the surveyed nursing homes (Wagner, McDonald, & Castle, 2012). The use of the t-Test was appropriate in this study since it allowed the comparison of the two groups, thus showing differences in the RSC ratings. This enabled accurate drawing of conclusion.

Statistical Assumptions

The use of appropriate data ensures that there is no violation of the assumptions of the choice of test used in a research (Plichta, & Kelvin, 2012). In the t-Test described and used above, the assumptions made included that the observations were independent. The researchers also assumed that the data they obtained in their research was normally distributed, and that it was interval or ration data (Plichta, & Kelvin, 2012). The populations used in the study were also assumed homogenous. Thus, they were expected to be similar in all other ways that were not subject to investigation by the study.

These assumptions were met in the study. The researchers used different ways of meeting them. Independence of the data was ensured through the analysis of institutions that were accredited and those that were not accredited, with the accreditation status providing the difference in the two sets of data. The populations were assumed homogenous in that they were both from the nursing homes. They had no other significant differences between them (Wagner, McDonald, & Castle, 2012). Therefore, the researchers met the assumptions of their study.

Levels of Measurement

The levels of measurement in statistical analysis developed by Stanley Smith Stevens arise from his theory of scale types (Plichta, & Kelvin, 2012). These include the “nominal, ordinal, interval and ratio types” (Plichta, & Kelvin, 2012, p. 5). The main scores used in the study by Wagner, McDonald, and Castle were the RSC scores for the institutions under study. The scores in the study were nominal measurements. The researchers obtained the means of these scores to compare the two values in order to make the necessary conclusions with regard to the study objectives.

The level of measurement used in this study was relevant in a number of ways. The nominal data allowed the obtaining of means in the population of institutions that were recognized and those that were not certified (Wagner, McDonald, & Castle, 2012). The measures were also appropriate in the study because they allowed researchers to make assumptions and the relevant inferences easily (Wagner, McDonald, & Castle, 2012). According to Plichta and Kelvin (2012), the levels of measurement used in any research should be chosen based on the particular characteristic to be examined in the study. These levels should be relevant to the study.

Data Displays

The methods of data display used in the presentation of data allow room for the simplification of the research findings. Methods of data presentation include bar charts, pie charts, tables, graphs, and histograms (Plichta, & Kelvin, 2012). Each of these methods has its own advantages. Each of them may be used in different types of research. In the research by Wagner, McDonald, and Castle (2012), tables are used as the only method of data display. The tables are two in number, with one displaying the score obtained from the research instrument that investigated the RSC in the different institutions while the second table shows the general characteristic of the respondents (Wagner, McDonald, & Castle, 2012).

The data display method used was appropriate for this study since tables were the best method of presenting the data obtained from this particular study. The strengths of data presentation using table include the relative ease with which readers can interpret the conclusions. Tables are also an efficient method of data presentation. They break monotony of presentation of data using plain text (Plichta, & Kelvin, 2012). The readers of the report can access the necessary information faster from the tables. Data presentation using tables also provides a visually appealing method where researchers can present their data effectively to their audience.

However, the method of data presentation has a number of weaknesses over other methods such as the use of graphs and charts. By using only one method of data presentation, the researchers were not able to create a good mental picture to their audience. The use of graphs, charts, and figures should have been a beater method of data presentation for this particular research. The researchers were also not able to use these tables to show the differences in the various values that they obtained from the research.


In conclusion, scholars need to guarantee that they choose the best methods of fact gathering, investigation, management, and explanation. This paper describes the use of the t-Test in evaluating the differences between two means. One of the research articles that used the test is described, with an analysis made in terms of how the researchers used the test to arrive at their conclusions. The study is also evaluated in terms of the assumptions, levels of measurement, and data presentation. These methods are appropriate. Hence, the conclusions made are credible.

Reference List

Plichta, B., & Kelvin, A. (2012). Munro’s statistical methods for health care research. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins.

Wagner, L., McDonald, S., & Castle, N. (2012). Relationship between Nursing Home Safety Culture and Joint Commission Accreditation. The Joint Commission Journal on Quality and Patient Safety, 38 (5), 207-215.

Cite this paper

Select style


NursingBird. (2022, April 13). The Non-Parametric or t-Tests Assessment. Retrieved from


NursingBird. (2022, April 13). The Non-Parametric or t-Tests Assessment.

Work Cited

"The Non-Parametric or t-Tests Assessment." NursingBird, 13 Apr. 2022,


NursingBird. (2022) 'The Non-Parametric or t-Tests Assessment'. 13 April.


NursingBird. 2022. "The Non-Parametric or t-Tests Assessment." April 13, 2022.

1. NursingBird. "The Non-Parametric or t-Tests Assessment." April 13, 2022.


NursingBird. "The Non-Parametric or t-Tests Assessment." April 13, 2022.