8210 wk5 discussion

Response 1

Leila Abouzaki 

RE: Discussion - Week 5

COLLAPSE

Top of Form

A null hypothesis (H0) is a statement that shows no difference between the population mean and some specified value which contradicts the research hypothesis (Frankfort-Nachmias et al., 2020). If the null hypothesis is true and the researcher rejects it nonetheless, it would be an incorrect decision or an error which is called a Type I error (Frankfort-Nachmias et al., 2020). Most users of statistics assume a value of .05 which represents a small risk of Type I error (Warner, 2012). However, in exploratory research, researchers sometimes use alpha levels (such as α= .10) that shows a higher risk of Type I error (Warner, 2012). When smaller alpha levels are used such as α= .01 or α= .001 it means researchers want to keep the theoretical risk of Type I error very small (Warner, 2012). Therefore, in the explorartory research, the .10 level shows a higher risk of Type I error rate which implies that a certain amount of tests will reject H0 (Magnusson, n.d.). 

The research paper claims a meaningful contribution to the literature based on finding statistically significant relationships between predictor and response variables. When the variable is measured in meaningful units, the difference can be interpreted as practical significance by the effect size that can be described as Cohen's d (Warner, 2012). As the research hypothesis (Ha) approaches H0 power will approach the alpha level (α) for small values of d (Magnusson, n.d.). Smaller values of practicial significance show less real world application or meaningfulness (Walden University, 2016). Thus, a study can be flawed and have uninformative results, and even when it is well designed and conducted, statistically significant outcomes can happen just by chance (Warner, 2012). Therefore, the results of the single study should never be treated as conclusive evidence (Warner, 2012).

References

Frankfort-Nachmias, C., Leon-Guerrero, A., & Davis, G. (2020). Social statistics for a diverse society (9th ed.). Sage Publications.

Magnusson, K. (n.d.). Welcome to Kristoffer Magnusson’s blog about R, Statistics, Psychology, Open Science, Data Visualization [blog].  http://rpsychologist.com/index.html

Walden University, LLC. (Producer). (2016). Meaningfulness vs. statistical significance [Video file]. Author.

Warner, R. M. (2012). Applied statistics from bivariate through multivariate techniques (2nd ed.). Sage Publications.

Bottom of Form

RESPONSE 2

Cherie McElroy-Burch 

RE: Discussion - Week 5

COLLAPSE

Top of Form

Gallo (2019) defines statistical significance as a tool to help quantify whether a result is likely due to probability or some other factor of interest. Statistical hypothesis testing is a procedure used to help evaluate hypotheses (answers to research questions) about population parameters based on sample statistics (Frankfort-Nachmias et al., 2020). According to Frankfort-Nachmias et al. (2020), the research hypothesis is a statement reflecting an expectation or prediction. The null hypothesis contradicts the research hypotheses and signifies no difference between specified populations. The p-value, resulting from a statistical test, tells researchers the probability associated with a particular set of observations. Research has shown that there are common misconceptions and misuse of the p-value. Ron Wasserstein, Executive Director of the American Statistics Association, asserts that the p-value was never planned to be used as a substitute for scientific reasoning (American Statistical Association, 2016). Wasserstein argues that the value of a single number (p-value) is not the total of a well-reasoned statistical argument. Ultimately, smaller p-values are valued over statistical and scientific reasoning (American Statistical Association, 2016). 

 

In the scenario, the researchers claimed a meaningful contribution to the literature based on finding statistically significant relationships between predictor and response variables. The researchers also noted that due to the exploratory research, a 95% significance level was not used; instead, the significance levels were relaxed to the .10 level. When confidence levels are decreased, the alpha increases and cause a higher possibility that the null hypothesis will be rejected. As a reviewer of this research, I would mention to the authors that just because a relationship is statistically significant between two variables does not mean that the relationship is considered meaningful or important theoretically (Frankfort-Nachmias et al., 2020). When evaluating research for meaningfulness, we look beyond the relationship of the two variables to determine if the relationship applies to the real world. In my opinion, this research is flawed and incomplete because it lacks a sample size, an explanation of meaningfulness, and an explanation of how the research can drive social change. More information is needed as a reviewer to determine if this research is truly meaningful.

 

References

 

A Refresher on Statistical Significance. (2021, December 9). Harvard Business Review. Retrieved June 28, 2022, from  https://hbr.org/2016/02/a-refresher-on-statistical-significance

 

Frankfort-Nachmias, C., Leon-Guerrero, A., & Davis, G. (2020). Social statistics for a diverse 

society (9th ed.). Sage Publications.

Bottom of Form

5