- Programs & Events
- Researchers' Toolkit
- News & Networking
"A picture is worth a thousand words." We have all experienced the struggle of trying to learn the "take-home message" by trying to filter through lots of numbers. We have also been left hanging on this message when there is not a sentence or two translating what these numbers mean. Authors are encouraged to enhance the readability of their findings by visually displaying the results of an interaction.
Read more about tables and figures.
TIP: Tables and figures should be used only when they are more efficient than text in presenting information (Don't just use them to be impressive)!
One component of many Results sections is a subheading of descriptive statistics such as measures of central tendency and variability for the entire sample, as well as correlations among variables to determine whether multicollinearity exists. In the event of very high correlations among selected variables, the authors should clearly spell out how this problem was handled (i.e., which variables were removed or combined).
Generally, a "group difference" research hypothesis is tested using some type of ANOVA procedure (e.g., ANOVA, MANOVA, repeated measures analysis), whereas a question of "relationships among variables" is tested with some forms of regression procedure (e.g., correlation, multiple regression, path analysis, structural equation modeling). Unnecessarily incorporating several different types of analyses without a clear sense of why these multiple analyses were necessary, focuses more on the number and types of statistical methods used than directly and unambiguously testing the original hypotheses.
Along with the required statistics denoting the test of significance (i.e., F value, degrees of freedom, p level), it is also crucial to follow up this standard of reporting with information such as: (a) which groups differed (main effect), or which groups differed at which trial blocks (interaction)? (b) in what direction are these differences? (c) what are the group means and standard deviations? (refer to a Table or Figure) and (d) what is the strength of the statistical significance (i.e., effect size)? In qualitative studies, ensure that data that represent the participants' voices are included to support the themes (meaning use direct quotes rather than summaries of what participants said).
Given that all relevant information is reported for a statistical procedure (e.g., factor analysis, canonical correlation), are the results interpreted accurately? What do these findings mean in nonstatistical terms? Does the visual of the interaction depict the associated verbal descriptions?