RES 6551 Module 4 APA Tables and Figures Example

Three Tables and Two Figures: Building the APA Displays for Chapter 4 of a Mediation Study of Leadership and Employee Voice

Student Name

American College of Education

RES6551: Analyzing the Dissertation Research

Module 4 Assignment

Instructor Name

November 2, 2026

Introduction

Readers of a results chapter often look at the tables before the text. A well-built table lets a committee member check a claim without hunting through paragraphs, and a well-built figure shows a pattern faster than any sentence. This paper designs the displays for Chapter 4 of my study of transformational leadership, psychological safety and voice among 221 hospital revenue cycle employees. It describes three tables and two figures in the order they will appear, gives the values each holds, explains the APA 7 conventions applied and lists the supporting material that will go to the appendices rather than the chapter.

Principles Applied

Four principles guided the design. Each display must stand alone, with a title, column labels and notes that define abbreviations and scales, so that it can be understood without the text (American Psychological Association, 2020). Displays and text divide the work: tables carry the full set of numbers and the text points to the ones that answer the research questions, without repeating every value. Intervals accompany estimates wherever possible, because they show precision in a way a p value alone does not (Cumming, 2014). And decimals are consistent: two places for descriptive statistics and correlations, three for regression coefficients and standard errors.

Table 1: Participant Characteristics

Table 1 has one column for each health system and a total column. Its rows show how many people, and what share, fell in each role and work arrangement, then summarize tenure with three numbers: average, spread and middle value. In the total column, roles read patient access 72 (32.6%), coding 54 (24.4%), denials and follow-up 48 (21.7%) and billing 47 (21.3%); work location reads on site 116 (52.5%) and remote or hybrid 105 (47.5%); tenure reads M = 4.95, SD = 3.64, Mdn = 3.6. Column totals are 100 for System A, 68 for System B and 53 for System C. Age and gender bands appear only in the total column, and a note explains that system-level demographic counts were withheld because some cells would have contained fewer than five people.

Table 2: Descriptive Statistics and Correlations

Table 2 lists the five study variables in rows: transformational leadership, psychological safety, voice, tenure and remote or hybrid work. Columns give the mean, standard deviation, observed range and Cronbach's alpha, followed by a correlation matrix below the diagonal with alphas on the diagonal in parentheses. The key cells read: leadership M = 3.48, SD = 0.81, range 1.43 to 5.00, alpha = .92; psychological safety M = 4.90, SD = 1.08, range 1.86 to 7.00, alpha = .92; voice M = 4.61, SD = 1.08, range 1.33 to 6.83, alpha = .91. Correlations are .59 between leadership and safety, .35 between leadership and voice and .44 between safety and voice. A note gives N = 221, states that correlations of .13 or larger in absolute value are significant at p < .05, defines remote work as coded 1 for remote or hybrid and gives each scale's response range.

Table 3: Regression Models for the Mediation Paths

Table 3 is the core display. It has three model columns, each with an unstandardized coefficient and its standard error in parentheses, and rows for the constant, leadership, psychological safety, tenure and remote work, followed by R-squared and F. Model 1 predicts voice from leadership and controls: leadership 0.468 (0.084), tenure -0.001 (0.019), remote 0.216 (0.136), R-squared = .132, F(3, 217) = 11.02. Model 2 predicts psychological safety: leadership 0.791 (0.073), tenure -0.026 (0.016), remote -0.012 (0.118), R-squared = .353, F(3, 217) = 39.45. Model 3 predicts voice with both: leadership 0.182 (0.099), psychological safety 0.362 (0.074), tenure 0.008 (0.018), remote 0.220 (0.130), R-squared = .218, F(4, 216) = 15.06. A final row beneath the models reports the indirect effect, 0.286, with its bootstrap standard error of 0.066 and 95% interval [0.162, 0.419]. Significance is marked with asterisks defined in the note, and 95% confidence intervals for the leadership and safety coefficients appear in a second line beneath each estimate.

Figure 1: The Mediation Model

Figure 1 is a path diagram with three boxes. Transformational leadership sits on the left, psychological safety at the top center and voice on the right. The arrow from leadership to safety is labeled a = 0.791**, the arrow from safety to voice b = 0.362*, and the arrow from leadership to voice carries two labels: c = 0.468** for the total effect and, in parentheses, c' = 0.182 for the direct effect. The diagram shows at a glance that the direct arrow weakens once the path through safety is included, which is the central finding of the study. Controls are listed in the figure note rather than drawn, to keep the diagram readable.

Figure 2: Bootstrap Distribution

Figure 2 is a histogram of the 5,000 bootstrap estimates of the indirect effect, with a vertical line at the point estimate of 0.286 and shaded tails beyond the 2.5th and 97.5th percentiles, 0.162 and 0.419. A dashed line marks zero, which lies well outside the shaded interval. The figure is included because some committee members are less familiar with bootstrapped intervals than with p values, and the histogram shows directly what it means for an interval to exclude zero (Hayes, 2022). Its shape, slightly skewed to the right, also illustrates why a percentile interval was preferred to one that assumes normality.

Supporting Evidence for the Appendices

Material that supports the results but would crowd the chapter goes to appendices. Appendix C holds the participant flow diagram from 241 started surveys to 221 cases. Appendix D holds the residual-versus-fitted plot, the normal Q-Q plot and the Breusch-Pagan result for the full model, together with the HC3 re-estimates. Appendix E holds the two sensitivity analyses, with system and role indicators added and with fast responders removed, laid out in the same three-model format as Table 3 so that estimates can be compared column by column. Appendix F holds item-level means, standard deviations and item-total correlations for each scale. Each appendix is referred to at the point in Chapter 4 where it becomes relevant.

Checking Displays Against Text

Before submission, every number in the Chapter 4 text will be checked against its table or figure, and every table value against the analysis output, using a printed checklist. Values will be copied from output files by script rather than typed, which removes most transcription errors. The check also confirms that each table and figure is mentioned in the text before it appears and that table numbers run in order. A colleague unfamiliar with the study will also be asked to read each table alone and say what it shows; any display that needs the text to be understood will be revised.

Conclusion

Three tables and two figures carry the results of Chapter 4: who took part, how the variables behaved and related, the three regression models with the indirect effect, the path model and the bootstrap distribution. Each display stands alone, follows APA 7 conventions and reports intervals with estimates. Supporting diagnostics and sensitivity results move to appendices, and a scripted check keeps text and displays consistent. Module 5 assembles the complete Chapter 4.

References

American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000

Cumming, G. (2014). The new statistics: Why and how. Psychological Science, 25(1), 7-29. https://doi.org/10.1177/0956797613504966

Hayes, A. F. (2022). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (3rd ed.). Guilford Press.