Research data · Sex, meta-analyses and retractions

The animals' sex, meta-analyses and retractions by test

Of the papers from 2021 to 2025 that PubMed tags with the sex of their animals, 24% used both sexes, up from 18% in 2011 to 2015.

The counts cover animal research papers that mention one of 90 behavioral tests in their title or abstract, as PubMed listed them on November 8, 2026.

See each test See retractions by test

of the papers from 2021 to 2025 with a sex tag used males only, and 10% used females only

of the papers with a sex tag that mention self-administration used both sexes in 2021 to 2025, up from 11% in 2011 to 2015

24 of the 5,392 papers from 2021 to 2025 that mention the Morris water maze are meta-analyses, the most of any test

of the 90 tests have at least one paper, from any year, that PubMed marks as retracted

Which sex the animals were

Among the subject headings PubMed gives a paper are the tags Male and Female, for the sex of the animals studied. Each bar splits the tagged papers into both sexes, males only and females only.

Papers that mention a test, in each period

Both sexes
Males only
Females only

2011 to 201518% both sexes, 72% males only and 10% females only, of 19,914 papers

2021 to 202524% both sexes, 66% males only and 10% females only, of 16,752 papers

These shares leave out papers whose subject headings carry no sex tag, and papers with no subject headings. Of the papers from 2021 to 2025 with subject headings, 9,084, or 35%, carried no sex tag. In 2011 to 2015 that share was half as large, so the rise from 18% to 24% using both sexes may not compare exactly.

Both sexes, test by test

Each bar is one test with 100 or more tagged papers from 2021 to 2025, ranked by the share that used both sexes. Each share pools every species. Of the 305 papers from 2021 to 2025 that mention the test ranked first, negative geotaxis and climbing assays, 201 used fruit flies and 112 have a sex tag.

Both sexes
Males only
Females only

The highest shares

Negative geotaxis and climbing assays65% both sexes, 24% males only and 11% females only, of 112 papers

Social interaction test47% both sexes, 46% males only and 7% females only, of 270 papers

Prepulse inhibition and acoustic startle47% both sexes, 44% males only and 10% females only, of 374 papers

Self-administration42% both sexes, 53% males only and 5% females only, of 886 papers

Three-chamber sociability test41% both sexes, 46% males only and 13% females only, of 201 papers

Marble burying41% both sexes, 46% males only and 13% females only, of 197 papers

Nest building41% both sexes, 25% males only and 34% females only, of 215 papers

Self-grooming40% both sexes, 38% males only and 22% females only, of 245 papers

The lowest shares

Catalepsy bar test11% both sexes, 85% males only and 3% females only, of 116 papers

Passive avoidance13% both sexes, 82% males only and 5% females only, of 614 papers

Y-maze13% both sexes, 79% males only and 8% females only, of 1,014 papers

Morris water maze13% both sexes, 79% males only and 7% females only, of 2,749 papers

Treadmill running14% both sexes, 74% males only and 13% females only, of 1,059 papers

Formalin test15% both sexes, 79% males only and 6% females only, of 261 papers

Tail suspension test16% both sexes, 72% males only and 12% females only, of 709 papers

Rotarod16% both sexes, 76% males only and 8% females only, of 930 papers

Show the 40 tests with 100 or more tagged papers

Sex of the animals by test
Test	Papers with a sex tag	Both sexes	Males only	Females only
Negative geotaxis and climbing assays	112	65%	24%	11%
Social interaction test	270	47%	46%	7%
Prepulse inhibition and acoustic startle	374	47%	44%	10%
Self-administration	886	42%	53%	5%
Three-chamber sociability test	201	41%	46%	13%
Marble burying	197	41%	46%	13%
Nest building	215	41%	25%	34%
Self-grooming	245	40%	38%	22%
Operant conditioning	496	40%	55%	5%
Light-dark box	340	35%	52%	13%
Active avoidance (shuttle box)	104	34%	60%	7%
Novelty-suppressed feeding	155	32%	56%	12%
Fear conditioning	1,057	31%	65%	5%
T-maze	198	28%	62%	10%
Wheel running	582	28%	53%	19%
Barnes maze	283	26%	65%	9%
Conditioned place preference	655	26%	70%	4%
Elevated plus maze	1,689	24%	64%	12%
Gait analysis	325	24%	60%	16%
Grip strength	696	21%	67%	12%
Open field	2,144	21%	68%	11%
Object location test	189	21%	69%	11%
Von Frey test	649	20%	70%	9%
Novel object recognition	1,663	20%	70%	10%
Tail flick	285	20%	74%	6%
Sucrose preference test	822	19%	71%	10%
Hot plate	361	18%	77%	5%
Forced swim test	1,590	18%	71%	11%
Radial arm maze	106	18%	75%	8%
Writhing test	250	18%	64%	19%
Wire hang and inverted grid	115	17%	75%	8%
Beam walking	243	17%	77%	5%
Rotarod	930	16%	76%	8%
Tail suspension test	709	16%	72%	12%
Formalin test	261	15%	79%	6%
Treadmill running	1,059	14%	74%	13%
Morris water maze	2,749	13%	79%	7%
Y-maze	1,014	13%	79%	8%
Passive avoidance	614	13%	82%	5%
Catalepsy bar test	116	11%	85%	3%

Where the share using both sexes rose or fell most

Each row is one of the 37 tests with 100 or more tagged papers in both periods. It shows the share that used both sexes in 2011 to 2015, then in 2021 to 2025.

2011 to 2015
2021 to 2025, a rise
2021 to 2025, a fall

The largest rises

Self-administration

11%42%

Operant conditioning

13%40%

Active avoidance (shuttle box)

18%34%

Marble burying

26%41%

Prepulse inhibition and acoustic startle

33%47%

Novelty-suppressed feeding

18%32%

The largest falls

61%41%

31%18%

32%21%

26%18%

24%16%

18%13%

Review articles and meta-analyses

A meta-analysis pools the results of many studies. These tests have the most meta-analyses among their papers from 2021 to 2025.

Reviews among the papers from 2021 to 2025, by test
Test	Meta-analyses	Systematic reviews	Reviews of any kind	All papers
Morris water maze	24	27	46	5,392
Forced swim test	16	21	46	2,988
Open field	13	15	29	3,980
Sucrose preference test	11	17	23	1,522
Elevated plus maze	11	13	29	2,883
Novel object recognition	9	11	26	3,019
Rotarod	9	8	19	1,844
Fear conditioning	8	14	101	2,010
Wheel running	6	11	31	992
Y-maze	6	9	19	1,988

One review can be both a systematic review and a meta-analysis, so the columns overlap.

Retracted papers

Each bar is one test and shows how many of its papers, from any year, PubMed marks as retracted. Among the 39 tests with 1,000 or more papers, the median is 1.2 retracted papers for every 1,000.

The tests with the most retracted papers

Morris water maze6.5 for every 1,000 of its 19,762 papers

129

Open field2.4 for every 1,000 of its 13,496 papers

Rotarod4.2 for every 1,000 of its 7,187 papers

Forced swim test2.1 for every 1,000 of its 12,343 papers

Novel object recognition3.1 for every 1,000 of its 8,027 papers

Y-maze4.1 for every 1,000 of its 5,413 papers

Sucrose preference test5.4 for every 1,000 of its 3,860 papers

Elevated plus maze1.6 for every 1,000 of its 11,716 papers

Passive avoidance2.0 for every 1,000 of its 8,839 papers

Treadmill running1.2 for every 1,000 of its 11,926 papers

How we count sex, meta-analyses and retractions

A paper tagged both Male and Female counts as using both sexes. The tags show which sexes a paper used, not whether it compared them.

Reviews come from PubMed's publication types Review, Systematic Review and Meta-Analysis, for papers from 2021 to 2025.

A retracted paper carries PubMed's publication type Retracted Publication, and the counts include papers from any year. They leave out the retraction notices themselves.

The file of all the numbers holds these counts for every test.

Download all the numbers