Applications of statistics in research

¶ … ethical issues that should be addressed by researchers in the field of psychology who use human subjects: the protection of the human participants and the protection and maintenance of the integrity of the research findings (Punch, 2013). This paper will discuss several important ethical considerations regarding the use of statistics in the integrity of research findings.

Perhaps the topic that is addressed most frequently regarding the ethical use of statistics and certainly is a current concern is the overreliance of statistical significance testing (Verdam, Oort, & Sprangers, 2014). The surprising to find that many researchers still believe that significant P. values indicate that the research results will replicate, smaller team P. values indicate a larger magnitudes and results, or that smaller P. values mean that the results are more important (Gorard, 2014; Punch, 2013). None of these are true. In fact, the American Psychological Association [APA] has suggested since the 1990s that effect sizes be reported along with the values for statistically significant results and this practice is become mandatory for many journals (Gorard, 2014). Nonetheless, many researchers continue to concentrate on hypothesis testing and statistically significant differences between means as opposed to more robust measures (Punch, 2013; Schneider, 2013).

A second common issue is ignoring the assumptions of statistical tests (Gorard, 2014; Punch, 2013). Nearly all general linear model procedures require that a certain set of statistical assumptions be met for use and these assumptions are often not considered by researchers. It is important to use procedures properly in order to avoid error and bias and statistical results. Moreover, when researchers find significant interactions remain effects and ANOVA designs they should use the proper post tests to compare means in order to determine where the interaction effects were made effects occur (Gorard, 2014; Punch, 2013).

An important consideration for any researcher is to make sure that the researcher does not make statements regarding statistical findings that are false or deceptive (Asendorpf, Conner, De Fruyt, De Houwer, Denissen et al., 2013; Gorard, 2014). The ethical researcher understands the implications of anti-statistical findings and reports these obligations accurately without speculating beyond the confines of the procedures used or of the methodology. Almost every small -- sample research study in psychological research has limited external validity and it is extremely important that researchers do not overstep the bounds created by their sampling method (Gorard, 2014; Punch, 2013).

The APA code of ethics clearly states that psychologist must recognize the limits of their compliments and scientific knowledge and make use of current scientific resources (Gorard, 2014). This is a principle that is most often connected with actual clinical practice; however, this is a guideline for researchers as well. Researchers should only use statistical techniques with which they are trained in unfamiliar with and should consultation and assistance when using procedures that they are not qualified to use. Researchers should only use new techniques only after appropriately being trained and supervised and their use (Punch, 2013). It is a mistake to believe that just because someone has performed research, had publications, and has an advanced degree that they are capable of utilizing and understanding any statistical method.