Recreating APA Manual Table 7.21 in R with apa7

Making tables in APA style (Part 21 of 24)

In this 24-part series, each of the tables in Chapter 7 of the Publication Manual of the American Psychological Association (7th Edition) is recreated with apa7, flextable, easystats, and tidyverse functions.

NoteHighlights

• Use of add_star_column
• Use of thin spaces   for footnote superscripts.

library(apa7)
library(flextable)
library(ftExtra)
library(tidyverse)
library(easystats)
library(lme4)
set_flextable_defaults(theme_fun = theme_apa, 
                       font.family = "Times New Roman")

Figure 1

Screenshot of the APA Manual’s Table 7.21

I did not have enough information to simulate the three models, so I input the data directly. This made it comparatively easy to recreate. Adding stars to the \chi^2 column was made easier by creating a temporary p-value column with values less than .001 and then applying add_star_column.

The APA Publication Manual recommends following the footnote symbols with a superscripted space. In the apa-note, the superscript symbols a, b, and c are followed by a thin space within the superscript. This gives a pleasing amount of separation.

```{r}
#| label: tbl-721
#| tbl-cap: "Results of Confirmatory Factor Analysis for the 
#|   Relationships Among Three Types of Intelligence"
#| apa-note: 
#|   - "Structural equation modeling was used for the analysis. 
#|     NFI = normed fit index; 
#|     CFI = comparative fit index; 
#|     RMSEA = root-mean-square error of 
#|     approximation."
#|   - "^a ^In Model A, all 57 items of social 
#|    intelligence, emotional intelligence, and cultural 
#|    intelligence were loaded onto one factor. ^b ^In
#|    Model B, the 21 items of social intelligence were 
#|    loaded onto one factor, and the 16 items of emotional 
#|    intelligence and the 20 items of cultural intelligence 
#|    were loaded onto another factor. ^c ^In Model C, 
#|    the 21 items of social intelligence were loaded onto one 
#|    factor, the 16 items of emotional intelligence were 
#|    loaded onto a second factor, and the 20 items of 
#|    cultural intelligence were loaded onto a third factor."
#|   - "^\\*\\*\\*^*p* < .001."
tibble(
  Model = c(
    "A: One-intelligence model^&thinsp;a^",
    "B: Two-intelligences mode^&thinsp;b^",
    "C: Three-intelligences model^&thinsp;c^"),
  Chi2 = c(10994.664, 10091.236, 8640.066),
  df = c(1539L, 1538L, 1536L),
  NFI = c(0.296, 0.354, 0.447),
  CFI = c(0.326, 0.39, 0.494),
  RMSEA = c(0.115, 0.109, 0.1),
  p = .0001
) |> 
  add_star_column(Chi2, merge = TRUE) |> 
  select(-p) |> 
  apa_flextable(column_formats = column_formats(
    accuracy = 0.001),
    layout = "fixed") |>
  width(width = c(2.15, 1.15, rep(3.2 / 4, 4)))
```

Table 1

Results of Confirmatory Factor Analysis for the Relationships Among Three Types of Intelligence

Model	χ2	df	NFI	CFI	RMSEA
A: One-intelligence model a	10994.664***	1,539	.296	.326	.115
B: Two-intelligences mode b	10091.236***	1,538	.354	.390	.109
C: Three-intelligences model c	8640.066***	1,536	.447	.494	.100

Note. Structural equation modeling was used for the analysis. NFI = normed fit index; CFI = comparative fit index; RMSEA = root-mean-square error of approximation.

^aIn Model A, all 57 items of social intelligence, emotional intelligence, and cultural intelligence were loaded onto one factor. ^bIn Model B, the 21 items of social intelligence were loaded onto one factor, and the 16 items of emotional intelligence and the 20 items of cultural intelligence were loaded onto another factor. ^cIn Model C, the 21 items of social intelligence were loaded onto one factor, the 16 items of emotional intelligence were loaded onto a second factor, and the 20 items of cultural intelligence were loaded onto a third factor.

^***p < .001.

• First
• Previous
• Next
• Last