Читать книгу The Research Experience - Ann Sloan Devlin - Страница 77

Null hypothesis: Hypothesis that there are no group differences or relationships between variables.

Null hypothesis significance testing procedure (NHSTP): Using statistical inference, a procedure for evaluating whether the null hypothesis should be rejected.

Alpha level: Probability (usually set at .05) of incorrectly rejecting the null hypothesis.

p value: Probability value based on the characteristics of the observed data used for hypothesis testing.

Confidence interval: Shows range of values that you can be sure contains the population mean a certain percentage of the time (e.g., 95%).

In most research that uses inferential statistics (and most research in the social and behavioral sciences does), the aim of the project is to be able to reject the null hypothesis of no difference between sampled populations. This process is sometimes called the null hypothesis significance testing procedure (NHSTP). As a scientific community, researchers therefore need a decision rule about when differences exist (that is, when to reject the null hypothesis represented by the data). Most members of the scientific community accept the standard of a statistical outcome with a probability value of .05 (5 / 100) or less frequent. The value .05, called the alpha level, represents the likelihood that in only 5 of 100 cases would the outcome occur by chance. This number is stated (acknowledged) before the analyses are conducted. This p value (or probability value) of .05 is the standard used by most journals for having significant results. In addition to the p value, it is common for journals to request that confidence intervals be provided. For example, a 95% confidence interval shows the range of values within which you can be sure the population mean is contained 95 out of 100 times. Current standards for APA journals are “estimates of appropriate effect sizes and confidence intervals” as a minimum, according to the APA manual (2020, p. 87).

It should be noted that at least one journal (Basic and Applied Social Psychology) has banned the use of NHSTP in papers submitted to the journal. In its place, the journal editors call for:

Strong descriptive statistics, including effect sizes. We also encourage the presentation of frequency or distributional data when this is feasible. Finally, we encourage the use of larger sample sizes than is typical in much psychology research, because as the sample size increases, descriptive statistics become increasingly stable and sampling error is less of a problem. (Trafimow & Marks, 2015, p. 1)

Most journals are interested in publishing results that advance the literature in some way. For that reason, research in which the null hypothesis is accepted (i.e., not rejected) is unlikely to be published.

In the most recent edition of the Publication Manual of the American Psychological Association (2020), an entire chapter is devoted to journal article reporting standards (JARS), covering standards for quantitative, qualitative, and mixed methods (i.e., containing both quantitative and qualitative) research. Additional resources beyond that chapter are provided by the APA Style JARS website (https://apastyle.apa.org/jars).