Reporting Descriptive and Frequency Analysis in APA Style
Based on Research With Fawad's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.
Use frequency analysis for categorical variables and report category counts (and basic percentages), while removing cumulative/valid cumulative percentage columns and any not-applicable categories.
Briefing
APA-style reporting for categorical and scale variables boils down to two workflows: generate the right frequency/descriptive outputs, then format and trim them to match APA conventions (especially what percentages to report and how tables are captioned).
For categorical variables such as gender, the process starts with a frequency table. In BlueSky Statistics (R-based), the frequency output for gender lists counts for each category (male coded as 1; female coded as 2), along with percentages and cumulative percentages. The key APA move is to remove columns that APA typically doesn’t require—specifically cumulative percentage and valid cumulative percentage—and also drop “not applicable” categories if they appear. After exporting the table (via Word/Excel/PDF or clipboard), the table is reformatted with APA-style borders: a bottom border across the table and a bottom border on the header row. The narrative write-up then pairs the table with a short description of the sample composition. In the example, the respondents are 414 males (53.49%) and 360 females (46.51%), for a total sample size of 774.
For descriptive statistics on numeric variables, the workflow shifts from frequency tables to summary statistics. Age is treated as a numerical variable, so the analysis reports the mean and standard deviation. The example mean age is 29.4525654 (reported with APA-typical rounding, commonly two decimals).
When the focus is a construct measured by multiple items—here, CSR measured through eight items—the reporting splits into an overall construct summary and item-level interpretation. The overall construct mean and standard deviation are computed from the transformed CSR variable. The example overall mean is 3.841 with a standard deviation of 0.671. Because the response scale runs from “strongly disagree” to “strongly agree,” the mean indicates a positive perception: respondents, on average, lean toward agreement rather than disagreement. The highest item mean is used to interpret what respondents most strongly endorsed; CSR 1 has the highest mean, and it corresponds to the hotel showing responsibility toward the environment.
Finally, the construct is reported by group—countries in this case (Pakistan, China, Italy). Instead of using the eight item scores directly, the analysis groups the overall CSR score by country and reports mean and standard deviation for each group. The example sample sizes are 312 for Pakistan, 278 for China, and 181 for Italy. The country-level CSR means are 3.58 (SD = 0.68) for Pakistan, 4.77 (SD = 0.58) for China, and 3.93 (SD = 0.62) for Italy. The resulting tables are then reformatted to APA standards (including removing unnecessary columns/rows and applying consistent borders), with captions such as “Table 2 Descriptive Statistics for CSR Constructor” in italics.
Cornell Notes
APA reporting for categorical and scale variables uses different outputs but a consistent table-and-text discipline. Gender is handled with a frequency table: report category counts (and typically percentages) while removing cumulative/valid cumulative percentage columns and any not-applicable categories. Age and other numeric variables use mean and standard deviation. For a multi-item construct like CSR (eight items), report an overall construct mean and SD, interpret the mean using the response scale direction, and optionally present item-level descriptive statistics (min, max, mean, SD). To compare CSR across groups, compute CSR’s overall score and run descriptive statistics by group (countries), reporting each country’s sample size, mean, and SD.
What columns from a frequency table should be removed to align with APA conventions, and why?
How does APA-style reporting differ between a categorical variable (gender) and a numerical variable (age)?
How should a multi-item construct like CSR be reported in APA style?
What does the overall CSR mean imply when the scale runs from strongly disagree to strongly agree?
How is CSR reported across countries, and what statistics are included?
What table formatting steps are emphasized for APA-style tables in the workflow?
Review Questions
- When using a frequency table for a categorical variable, which percentage columns are typically excluded in APA-style reporting?
- What statistics should be reported for a multi-item construct’s overall score versus its individual items?
- How do you report a construct (CSR) by group (countries) without listing all item scores?
Key Points
- 1
Use frequency analysis for categorical variables and report category counts (and basic percentages), while removing cumulative/valid cumulative percentage columns and any not-applicable categories.
- 2
Format APA-style tables with consistent borders—especially a bottom border across the table and a bottom border on the header row.
- 3
For numerical variables like age, report mean and standard deviation rather than frequencies.
- 4
For multi-item constructs (CSR), report an overall construct mean and standard deviation, then interpret the mean direction using the response scale anchors.
- 5
Identify the highest-scoring construct item (e.g., CSR 1) and translate its mean into a substantive statement tied to what that item measures.
- 6
When comparing constructs across groups (countries), group the overall construct score and report each group’s sample size, mean, and standard deviation.