grimmer_map_total_n {scrutiny} | R Documentation |
GRIMMER-testing with hypothetical group sizes
Description
When reporting group means, some published studies only report the total sample size but no group sizes corresponding to each mean. However, group sizes are crucial for GRIMMER-testing.
In the two-groups case, grimmer_map_total_n()
helps in these ways:
It creates hypothetical group sizes. With an even total sample size, it incrementally moves up and down from half the total sample size. For example, with a total sample size of 40, it starts at 20, goes on to 19 and 21, then to 18 and 22, etc. With odd sample sizes, it starts from the two integers around half.
It GRIMMER-tests all of these values together with the group means.
It reports all the scenarios in which both "dispersed" hypothetical group sizes are GRIMMER-consistent with the group means.
All of this works with one or more total sample sizes at a time. Call
audit_total_n()
for summary statistics.
Usage
grimmer_map_total_n(
data,
x1 = NULL,
x2 = NULL,
sd1 = NULL,
sd2 = NULL,
dispersion = 0:5,
n_min = 1L,
n_max = NULL,
constant = NULL,
constant_index = NULL,
...
)
Arguments
data |
Data frame with string columns |
x1 , x2 , sd1 , sd2 |
Optionally, specify these arguments as column names in
|
dispersion |
Numeric. Steps up and down from half the |
n_min |
Numeric. Minimal group size. Default is 1. |
n_max |
Numeric. Maximal group size. Default is |
constant |
Optionally, add a length-2 vector or a list of length-2
vectors (such as a data frame with exactly two rows) to accompany the pairs
of dispersed values. Default is |
constant_index |
Integer (length 1). Index of |
... |
Arguments passed down to |
Value
A tibble with these columns:
-
x
, the group-wise reported input statistic, is repeated in row pairs. -
n
is dispersed from half the inputn
, withn_change
tracking the differences. -
both_consistent
flags scenarios where both reportedx
values are consistent with the hypotheticaln
values. -
case
corresponds to the row numbers of the input data frame. -
dir
is"forth"
in the first half of rows and"back"
in the second half."forth"
means thatx2
from the input is paired with the larger dispersedn
, whereas"back"
means thatx1
is paired with the larger dispersedn
. Other columns from
grimmer_map()
are preserved.
Summaries with audit_total_n()
You can call
audit_total_n()
following up on grimmer_map_total_n()
to get a tibble with summary statistics. It will have these columns:
-
x1
,x2
,sd1
,sd2
, andn
are the original inputs. -
hits_total
is the number of scenarios in which all ofx1
,x2
,sd1
, andsd2
are GRIMMER-consistent. It is the sum ofhits_forth
andhits_back
below. -
hits_forth
is the number of both-consistent cases that result from pairingx2
andsd2
with the larger dispersedn
value. -
hits_back
is the same, exceptx1
andsd1
are paired with the larger dispersedn
value. -
scenarios_total
is the total number of test scenarios, whether or not bothx1
andsd1
as well asx2
andsd2
are GRIMMER-consistent. -
hit_rate
is the ratio ofhits_total
toscenarios_total
.
References
Bauer, P. J., & Francis, G. (2021). Expression of Concern: Is It Light or Dark? Recalling Moral Behavior Changes Perception of Brightness. Psychological Science, 32(12), 2042–2043. https://journals.sagepub.com/doi/10.1177/09567976211058727
Allard, A. (2018). Analytic-GRIMMER: a new way of testing the possibility of standard deviations. https://aurelienallard.netlify.app/post/anaytic-grimmer-possibility-standard-deviations/
Bauer, P. J., & Francis, G. (2021). Expression of Concern: Is It Light or Dark? Recalling Moral Behavior Changes Perception of Brightness. Psychological Science, 32(12), 2042–2043. https://journals.sagepub.com/doi/10.1177/09567976211058727
See Also
function_map_total_n()
, which created the present function using
grimmer_map()
.
Examples
# Run `grimmer_map_total_n()` on data like these:
df <- tibble::tribble(
~x1, ~x2, ~sd1, ~sd2, ~n,
"3.43", "5.28", "1.09", "2.12", 70,
"2.97", "4.42", "0.43", "1.65", 65
)
df
grimmer_map_total_n(df)
# `audit_total_n()` summaries can be more important than
# the detailed results themselves.
# The `hits_total` column shows all scenarios in
# which both divergent `n` values are GRIMMER-consistent
# with the `x*` values when paired with them both ways:
df %>%
grimmer_map_total_n() %>%
audit_total_n()
# By default (`dispersion = 0:5`), the function goes
# five steps up and down from `n`. If this sequence
# gets longer, the number of hits tends to increase:
df %>%
grimmer_map_total_n(dispersion = 0:10) %>%
audit_total_n()