error bars on plots

idhgt

ggplot

Author

Published

April 25, 2023

Repurposing this APA figures post as a IDHTG (I don’t have to google) post.

As I write my first paper reporting data analysis coming out of R (woot!!!), here are some notes summarising all the googling I have done this morning about how to produce APA style figures in ggplot.

Load libraries

Start by loading tidyverse to get ggplot, here to make finding the data easy, and papaja to get the theme_apa() function.

library(tidyverse)

── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.4     ✔ readr     2.1.5
✔ forcats   1.0.0     ✔ stringr   1.5.1
✔ ggplot2   3.5.1     ✔ tibble    3.2.1
✔ lubridate 1.9.4     ✔ tidyr     1.3.1
✔ purrr     1.0.4     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

library(here)

here() starts at /home/runner/work/jenrichmond.github.io/jenrichmond.github.io

library(papaja)

Loading required package: tinylabels

Read in data

plotdata <- read_csv("plotdata.csv")

New names:
Rows: 8 Columns: 9
── Column specification
──────────────────────────────────────────────────────── Delimiter: "," chr
(4): direction, group, detailtype, groupnew dbl (5): ...1, mean, stdev, n,
stderr
ℹ Use `spec()` to retrieve the full column specification for this data. ℹ
Specify the column types or set `show_col_types = FALSE` to quiet this message.
• `` -> `...1`

head(plotdata)

# A tibble: 6 × 9
   ...1 direction group     detailtype  mean stdev     n stderr groupnew       
  <dbl> <chr>     <chr>     <chr>      <dbl> <dbl> <dbl>  <dbl> <chr>          
1     1 future    control   episodic    9.46  4.04    28  0.764 control group  
2     2 future    control   semantic    4.57  2.35    28  0.444 control group  
3     3 future    induction episodic    9.38  3.62    29  0.672 induction group
4     4 future    induction semantic    4.69  2.85    29  0.530 induction group
5     5 past      control   episodic   11.2   6.67    28  1.26  control group  
6     6 past      control   semantic    5.5   5.53    28  1.05  control group

Basic ggplot (columns)

Plot separate bars for episodic vs semantic details, by past and future events, separately for kids in the control group vs. induction group. Get pairs of columns using position = “dodge”.

plotdata %>%
  ggplot(aes(x= detailtype, y = mean, fill = direction)) +
    geom_col(position = "dodge") +
  facet_wrap(~ groupnew)

Add error bars

plotdata %>%
  ggplot(aes(x= detailtype, y = mean, fill = direction)) +
    geom_col(position = "dodge") +
  facet_wrap(~ groupnew) + geom_errorbar(aes(ymin=mean-stderr, ymax=mean+stderr),
                  size=.3,    # Thinner lines
                    width=.2,
                      position=position_dodge(.9))

Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` instead.

APA-ise

The theme_apa() from the pajaja package does most of the APAising. Gets rid of the grey and gridlines. But for some reason, now the bars are floating.

plotdata %>%
  ggplot(aes(x= detailtype, y = mean, fill = direction)) +
    geom_col(position = "dodge") +
  facet_wrap(~ groupnew) + geom_errorbar(aes(ymin=mean-stderr, ymax=mean+stderr),
                  size=.3,    # Thinner lines
                    width=.2,
                      position=position_dodge(.9)) +
  theme_apa(base_size = 14)

Fix x and y axis

Extend y axis scale and make the bars sit on the x axis

plotdata %>%
  ggplot(aes(x= detailtype, y = mean, fill = direction)) +
    geom_col(position = "dodge") +
  facet_wrap(~ groupnew) + geom_errorbar(aes(ymin=mean-stderr, ymax=mean+stderr),
                  size=.3,    # Thinner lines
                    width=.2,
                      position=position_dodge(.9)) +
  theme_apa(base_size = 14) +
  scale_y_continuous(expand = c(0, 0), limits = c(0, 15)) # expand 0,0 to make the bars sit down

Fix axis labels

Use the \n notation to break a label or title across two lines

plotdata %>%
  ggplot(aes(x= detailtype, y = mean, fill = direction)) +
    geom_col(position = "dodge") +
  facet_wrap(~ groupnew) + geom_errorbar(aes(ymin=mean-stderr, ymax=mean+stderr),
                  size=.3,    # Thinner lines
                    width=.2,
                      position=position_dodge(.9)) +
  theme_apa(base_size = 14) +
  scale_y_continuous(expand = c(0, 0), limits = c(0, 15)) +
   labs(x="Detail type", y="Mean number of details \n produced")

Make grey scale

Use scale_fill_grey(), values 1 = white and 0 = black, specify values in between to get shades of grey

plotdata %>%
  ggplot(aes(x= detailtype, y = mean, fill = direction)) +
    geom_col(position = "dodge") +
  facet_wrap(~ groupnew) + geom_errorbar(aes(ymin=mean-stderr, ymax=mean+stderr),
                  size=.3,    # Thinner lines
                    width=.2,
                      position=position_dodge(.9)) +
  theme_apa(base_size = 14) +
  scale_y_continuous(expand = c(0, 0), limits = c(0, 15)) +
   labs(x="Detail type", y="Mean number of details \n produced") +
  scale_fill_grey(start = 0.40, end = 0.6)

Save as png to add to your paper

Use ggsave(“nameoffile.png”) to save the last plot as png.

ggsave("featured.png")

Saving 7 x 5 in image