Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 2.0

In the paper

“Studies measuring the impact of ARRIVE on the quality of reporting have produced mixed results [6–11]”

What the sources say

• Supported

  In “Are the Guidelines Being Ignored?” (page unavailable), the authors compare reporting before and after ARRIVE endorsement. They find increased reporting of animal characteristics, but little change in blinding, rare reporting of randomisation and power calculations, and less appropriate statistical analysis on one measure. This study therefore supports the cited sentence as an example of mixed results, without establishing what every study in the citation range found. (LLM, conf=0.87)

  cites [6] D Baker, 2014, Two years later: journals are not yet enforcing the ARRIVE guidelines on reporting standards for pre-clinical animal studies, PLoS Biol source
• Supported

  The Methods (Statistics) compare reporting before and after ARRIVE, so this is a study measuring its impact. The paper reports that ARRIVE’s publication did not seem to improve reporting quality. As one of several studies cited for mixed results, it supports the no-improvement side of that collective statement. (LLM, conf=0.88)

  cites [7] JE Gulin, 2015, Quality of reporting and adherence to ARRIVE guidelines in animal studies for Chagas disease preclinical drug research: a systematic review, PLoS Negl Trop Dis source
• Partly supported

  The Results (page unavailable) report a statistically significant but small increase in reporting after ARRIVE was published; fewer than half of the assessed items were still reported. The authors also say extraneous factors may explain the increase. This study contributes a qualified finding, but it does not establish that studies of ARRIVE’s impact collectively produced mixed results. (LLM, conf=0.82)

  cites [8] MT Avey, 2016, The devil is in the details: incomplete reporting in preclinical animal research, PLoS ONE sourceExperimental — accuracy not yet measured
• Partly supported

  In the Discussion (page unavailable), researchers who knew the ARRIVE guidelines scored higher on a self-reported reporting measure. The Introduction and Discussion also say reporting remained low after journal endorsement. These contrasting observations fit the idea of mixed results, but this survey did not directly measure ARRIVE’s impact on the quality of published reporting. (LLM, conf=0.82)

  cites [9] TS Reichlin, 2016, The researchers’ view of scientific rigor—survey on the conduct and reporting of in vivo research, PLoS ONE sourceExperimental — accuracy not yet measured
• Supported

  The Abstract and Discussion report that this study found no meaningful improvement in reporting quality from journal support for ARRIVE. As one of several studies cited for a collective statement about mixed results, it supports the no-improvement side of that statement; it need not establish the positive side on its own. (LLM, conf=0.88)

  cites [10] V Leung, 2018, ARRIVE has not ARRIVEd: Support for the ARRIVE (Animal Research: Reporting of in vivo Experiments) guidelines does not improve the reporting quality of papers in animal welfare, analgesia or anesthesia, PLoS ONE source
• Supported

  The Results report no improvement in full ARRIVE compliance, while the Discussion reports improvement in one individual ARRIVE reporting item. The Discussion also contrasts this trial’s findings with reports of improved reporting after a different, mandated checklist intervention. Together, these passages support the broad statement that reporting-quality findings have been mixed. (LLM, conf=0.86)

  cites [11] K Hair, 2019, A randomised controlled trial of an intervention to improve compliance with the ARRIVE guidelines (IICARus), Research Integrity and Peer Review source

In the paper

“Low power has three effects: first, within the experiment, real effects are more likely to be missed; second, when an effect is detected, this will often be an overestimation of the true effect size [24]; and finally, when low power is combined with publication bias, there is an increase in the false positive rate in the published literature [36].”

What the sources say

• ?Could not verify

  The section ‘What is the extent of the problem?’ discusses how correcting pseudoreplication can leave an experiment with low statistical power. The Introduction discusses inflated apparent evidence from pseudoreplication. Neither passage addresses whether effects detected in low-power studies tend to overestimate the true effect size; searches for effect-size overestimation and significant results in small samples found no such passage in our copy. (LLM, conf=0.76)

  cites [24] SE Lazic, 2018, What exactly is ‘N’ in cell culture and animal experiments?, PLoS Biol sourceExperimental — accuracy not yet measured

In the paper

“Low power has three effects: first, within the experiment, real effects are more likely to be missed; second, when an effect is detected, this will often be an overestimation of the true effect size [24]; and finally, when low power is combined with publication bias, there is an increase in the false positive rate in the published literature [36].”

What the sources say

• Supported

  The Abstract (page not supplied) says lower power reduces the probability that a significant finding reflects a true effect and that publishing biases may leave the literature heavily contaminated by false positives. “Low power in the presence of other biases” (page not supplied) adds that publication bias is more likely to affect underpowered studies. Together, these passages support the cited clause. (LLM, conf=0.88)

  cites [36] KS Button, 2013, Power failure: why small sample size undermines the reliability of neuroscience, Nat Rev Neurosci source

In the paper

“Conflating experimental units with subsamples or repeated measurements can lead to artificial inflation of the sample size.”

What the sources say

• Supported

  The Background defines sample size as the number of experimental units and says that multiple observations from one rat cannot be summed to give a larger sample size. Its pseudoreplication example illustrates the error by treating 20 measurements from two rats as 20 independent data points. (LLM, conf=0.94)

  cites [26] SE Lazic, 2010, The problem of pseudoreplication in neuroscientific studies: is it affecting your analysis?, BMC Neurosci source

In the paper

“Animals, experimental units, or data points that are unaccounted for can lead to instances in which conclusions cannot be supported by the raw data [52].”

What the sources say

• Partly supported

  The section “Reproducibility and replicability in general science: a crisis?” (page unavailable) says that some studies are difficult to reproduce because their figures, tables, and conclusions cannot be regenerated from the study data. It does not identify unaccounted-for animals, experimental units, or data points as the cause, so that specific part of the claim is unsupported. (LLM, conf=0.82)

  cites [52] N Kafkafi, 2018, Reproducibility and replicability of rodent phenotyping in preclinical studies, Neurosci Biobehav Rev sourceExperimental — accuracy not yet measured

In the paper

“Systematic reviews have shown that animal experiments that do not report randomisation or other bias-reducing measures such as blinding are more likely to report exaggerated effects that meet conventional measures of statistical significance [67–69].”

What the sources say

• Partly supported

  The Results (Randomization; page not available) report larger effect sizes in studies without randomization, and the Discussion says failure to randomize likely overestimates treatment benefits. But the Results (Blinding; page not available) find no overall difference for blinding in the primary analysis. The paper reports statistical significance for some pooled comparisons; it does not show that studies omitting these measures are more likely to report individually statistically significant effects. (LLM, conf=0.88)

  cites [67] JA Hirst, 2014, The need for randomization in animal trials: an overview of systematic reviews, PLoS ONE sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [68] HM Vesterinen, 2010, Improving the translational hit of experimental treatments in multiple sclerosis, Multiple Sclerosis Journal source
• Partly supported

  The Results report that animal-study abstracts reporting neither randomization nor blinding were more likely to report a positive outcome than those reporting both methods. The paper measured whether a difference met each author's definition of significance, not whether effect sizes were exaggerated or met a uniform conventional threshold. Its Limitations also restrict the analysis to abstracts from one emergency medicine meeting, so it does not establish the broader claim about animal experiments generally. (LLM, conf=0.87)

  cites [69] V Bebarta, 2003, Emergency medicine animal research: does use of randomization and blinding affect the results?, Acad Emerg Med sourceExperimental — accuracy not yet measured

In the paper

“Failure to disclose all the outcomes that were measured introduces bias in the literature, as positive outcomes (e.g., those statistically significant) are reported more often [88–91].”

What the sources say

• ?Not checked — no full text of the sourcecites [88] LK John, 2012, Measuring the prevalence of questionable research practices with incentives for truth telling, Psychol Sci source
• Supported

  The Introduction describes selective non-reporting of outcomes within published studies as outcome reporting bias and says statistically significant results have higher odds of being fully reported. The Discussion likewise reports that statistically significant outcomes have higher odds of being fully reported. (LLM, conf=0.93)

  cites [89] K Dwan, 2008, Systematic review of the empirical evidence of study publication bias and outcome reporting bias, PLoS ONE source
• Supported

  The Introduction says selective outcome reporting can involve measuring many outcomes but reporting only those with the best results, leaving an inflated proportion of statistically significant findings in the published evidence. The Discussion also describes studies potentially reporting only a few findings guided by their statistical significance. (LLM, conf=0.93)

  cites [90] KK Tsilidis, 2013, Evaluation of excess significance bias in animal studies of neurological diseases, PLoS Biol source
• Partly supported

  The Discussion reports that positive animal-stroke studies are more likely to be published than neutral ones, biasing the literature. But the paper examines missing experiments, not whether researchers selectively disclose positive outcomes among all outcomes measured in a study. (LLM, conf=0.88)

  cites [91] ES Sena, 2010, Publication bias in reports of animal stroke studies leads to major overstatement of efficacy, PLoS Biol sourceExperimental — accuracy not yet measured

In the paper

“There are, however, compelling data from systematic reviews showing that nonblinded outcome assessment leads to the treatment effects being overestimated, and the lack of bias-reducing measures such as randomisation and blinding can contribute to as much as 30%–45% inflation of effect sizes [67,68,85].”

What the sources say

• Partly supported

  The Introduction (Bias in Animal Studies) reports a human-trial study finding exaggerated odds ratios when outcome assessment was not blinded, and the Results (3. Blinding) find an effect for neurobehavioral outcomes. But this paper’s primary analysis found no significant overall difference between blinded and nonblinded animal studies, so it does not support the claim as an unqualified finding. (LLM, conf=0.86)

  cites [67] JA Hirst, 2014, The need for randomization in animal trials: an overview of systematic reviews, PLoS ONE sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [68] HM Vesterinen, 2010, Improving the translational hit of experimental treatments in multiple sclerosis, Multiple Sclerosis Journal source
• Partly supported

  The Discussion says efficacy may have been overstated by methodological flaws and identifies blinded outcome assessment as a measure that might reduce bias. But Potential Weaknesses says the impact of blinded outcome assessment is not yet clear, so the source does not establish the claim’s categorical causal conclusion. The Results report lower efficacy estimates with randomization and allocation concealment, but do not substantiate the cited sentence’s 30%–45% inflation figure. (LLM, conf=0.86)

  cites [85] MR Macleod, 2008, Evidence for the efficacy of NXY-059 in experimental focal cerebral ischaemia is confounded by study quality, Stroke sourceExperimental — accuracy not yet measured

In the paper

“Studies that claim to test a hypothesis but do not specify a predefined primary outcome measure or those that change the primary outcome measure after data were collected (also known as primary outcome switching) are liable to selectively report only statistically significant results, favouring more positive findings [94].”

What the sources say

• Supported

  The Introduction identifies choosing which response variables to report after analysis as a p-hacking practice and explains that selective reporting favors statistically significant findings. The paper also recommends prespecifying methods before data collection so readers can assess that risk. It does not use the term “primary outcome switching,” but the cited sentence is a reasonable paraphrase of the risk it describes. (LLM, conf=0.87)

  cites [94] ML Head, 2015, The extent and consequences of p-hacking in science, PLoS Biol source

In the paper

“Inferential statistics based on nonrandomised group allocation are not valid [65,66].”

What the sources say

• ?Not checked — no full text of the sourcecites [65] BG Greenberg, 1951, Why randomize?, Biometrics source
• Partly supported

  Altman and Bland explain that random allocation lets treatment-group differences be assessed using statistical theory about random samples (p. 1209). They also describe bias risks in nonrandomised comparisons, but do not say that all inferential statistics based on nonrandomised allocation are invalid; they acknowledge situations where randomisation is not possible. (LLM, conf=0.85)

  cites [66] DG Altman, 1999, Statistics notes. Treatment allocation in controlled trials: why randomise?, BMJ sourceExperimental — accuracy not yet measured

In the paper

“There are, however, compelling data from systematic reviews showing that nonblinded outcome assessment leads to the treatment effects being overestimated, and the lack of bias-reducing measures such as randomisation and blinding can contribute to as much as 30%–45% inflation of effect sizes [67,68,85].”

What the sources say

• Partly supported

  The Introduction (Bias in Animal Studies; page unavailable) reports a 30% exaggeration of odds ratios in human trials lacking allocation concealment. The Conclusions report exaggerated effect sizes in animal studies lacking randomization, but the paper does not report 30%–45% inflation from lack of randomization and blinding; the Results (Blinding; page unavailable) found no significant overall difference in its primary analysis. (LLM, conf=0.88)

  cites [67] JA Hirst, 2014, The need for randomization in animal trials: an overview of systematic reviews, PLoS ONE sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [68] HM Vesterinen, 2010, Improving the translational hit of experimental treatments in multiple sclerosis, Multiple Sclerosis Journal source
• Partly supported

  The Results report higher estimated NXY-059 efficacy in animal studies that did not report randomization or allocation concealment, supporting the general concern about bias. They do not report 30%–45% inflation of effect sizes, and they found no significant effect of reported blinded outcome assessment. The Discussion describes methodological flaws as a possible reason efficacy was overstated. (LLM, conf=0.88)

  cites [85] MR Macleod, 2008, Evidence for the efficacy of NXY-059 in experimental focal cerebral ischaemia is confounded by study quality, Stroke sourceExperimental — accuracy not yet measured

In the paper

“For most species, precise reporting of age is more informative than a description of the developmental status (e.g., a mouse referred to as an adult can vary in age from 6 to 20 weeks [108]).”

What the sources say

• Partly supported

  The Discussion (‘Age considered “adult”’) recommends reporting actual age instead of a life-stage term, and the Results show that ‘adult’ covered 6–20 weeks for mice. The paper examines rodents, however, so it does not establish the claim for ‘most species.’ (LLM, conf=0.88)

  cites [108] SJ Jackson, 2017, Does age matter? The impact of rodent age on study outcomes, Lab Anim sourceExperimental — accuracy not yet measured

In the paper

“For example, nocturnal animals may not show behavioural signs of discomfort during the day [134].”

What the sources say

• Supported

  Section 2.5.5 (Assessment timing; page not provided) recommends observing animals when they are normally active to reduce the chance of missing essential signs. It notes that most rodents are predominantly active during the dark period, supporting the claim that daytime observation may miss behavioural signs of discomfort. (LLM, conf=0.89)

  cites [134] P Hawkins, 2011, A guide to defining and implementing protocols for the welfare assessment of laboratory animals: eleventh report of the BVAAWF/FRAME/RSPCA/UFAW Joint Working Group on Refinement, Lab Anim source

In the paper

“Protocols vary depending on species, strain, and outcome; for example, physiological acclimatisation following transportation of different animals can take anywhere from 24 hours to more than 1 week [139].”

What the sources say

• Supported

  In “Acclimatization Following Transportation” (page not provided), the paper says primary stress mediators generally return to normal within 24 hours, while reproductive measures can take longer than one week. It also describes variation by animal age and genotype, supporting the citing sentence’s point that acclimatization time varies. (LLM, conf=0.93)

  cites [139] JA Obernier, 2006, Establishing an appropriate period of acclimatization following transportation of laboratory animals, ILAR Journal source

In the paper

“Procedural acclimatisation immediately before a procedure allows stabilisation of the animals’ responses after unaccustomed handling, novel environments, and previous procedures, which otherwise can induce behavioural and physiological changes [140,141].”

What the sources say

• Partly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites [140] DD Krahn, 1990, The anorectic effects of CRH and restraint stress decrease with repeated exposures, Biol Psychiatry sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [141] DL Pitman, 1988, Plasma corticosterone levels during repeated presentation of two intensities of restraint stress: chronic stress and habituation, Physiol Behav source

In the paper

“Along with innate circadian rhythms, these can affect research outcomes such as behavioural, physiological, and immunological parameters [132,133].”

What the sources say

• Supported

  The Discussion states that the behavioural, physiological, and immunological effects of repeated psychosocial stress depend strongly on the time of day of exposure, highlighting the circadian clock’s importance. This supports the claim that timing and circadian rhythms can affect these measured outcomes, in the study’s male-mouse stress context. (LLM, conf=0.92)

  cites [132] MS Bartlang, 2012, Time matters: pathological effects of repeated psychosocial stress during the active, but not inactive, phase of male mice, J Endocrinol source
• ?Not checked — no full text of the sourcecites [133] AK Paul, 2017, Morphine dosing strategy plays a key role in the generation and duration of the produced antinociceptive tolerance, Neuropharmacology source

In the paper

“The species, strain, substrain, sex, weight, and age of animals are critical factors that can influence most experimental results [103–107].”

What the sources say

• Partly supported

  The passage on animal-model selection (Abstract; page not provided) says that sex can affect the outcome being measured, supporting sex as a relevant experimental factor. The paper does not establish that species, strain, substrain, weight, and age influence results, or that these factors influence most experimental results. (LLM, conf=0.88)

  cites [103] JA Clayton, 2014, Policy: NIH to balance sex in cell and animal studies, Nature News sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [104] S Shapira, 2002, Aging has a complex effect on a rat model of ischemic stroke, Brain Res source
• Partly supported

  Results §3.3 (page not available) identifies age as a significant factor in murine gut microbiome differences, and Results §3.1 reports greater allergic airway inflammation in old mice. The paper does not establish that species, strain, substrain, sex, or weight influence results, or that these factors affect most experimental results. (LLM, conf=0.85)

  cites [105] M Vital, 2015, Alterations of the murine gut microbiome with age and allergic airway disease, J Immunol Res sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [106] JA Bouwknecht, 2002, Behavioral and physiological mouse assays for anxiety: a survey in nine mouse strains, Behav Brain Res source
• Partly supported

  The Discussion reports phenotypic differences between C57BL/6J and C57BL/6N mice that may affect comparisons of experimental results, and the Results describe some differences by sex. The paper does not establish that species, weight, or age influence results, or that the listed factors influence most experimental results. The indexed passages have no page numbers. (LLM, conf=0.87)

  cites [107] MM Simon, 2013, A comparative phenotypic and genomic analysis of C57BL/6J and C57BL/6N mouse strains, Genome Biol sourceExperimental — accuracy not yet measured

In the paper

“Cage/tank/housing system (type and dimensions) | Affects behaviour [177] and fear learning [178].”

What the sources say

• Partly supported

  The Discussion reports increased aggression with larger group sizes and floor areas in male BALB mice, so the paper supports an effect on that specific behaviour. It also says space allowance had little effect on the other behaviours studied; the Results sections report no differences in home-cage activity or open-field behaviour across cage conditions. The unqualified claim that housing type and dimensions affect behaviour overstates this limited finding. Fear learning is attributed to a different citation. (LLM, conf=0.85)

  cites [177] JD Bailoo, 2018, Evaluation of the effects of space allowance on measures of animal welfare in laboratory mice, Sci Rep sourceExperimental — accuracy not yet measured

In the paper

“Food and water (type, composition, supplier, and access) | Affects body weight, tumour development, nephropathy severity [181], and the threshold for developing parkinsonian symptoms [182].”

What the sources say

• Partly supported

  The Abstract concludes that a high-fat diet may lower the threshold for developing Parkinson’s disease, but only in the context of dopamine-specific toxin exposure in rats. It reports greater dopamine depletion, not earlier parkinsonian symptoms: locomotor activity did not differ between diet groups. The paper does not establish effects of water, supplier, or access. (LLM, conf=0.88)

  cites [182] JK Morris, 2010, Neurodegeneration in an animal model of Parkinson’s disease is exacerbated by a high-fat diet, American Journal of Physiology-Regulatory, Integrative and Comparative Physiology sourceExperimental — accuracy not yet measured

In the paper

“Bedding and nesting material | Affects behavioural responses to stress [184] and pain [185].”

What the sources say

• Partly supported

  In “Behavioral control over temperature” (page unavailable), the paper describes nest building as a behavioral way for mice to control heat loss, and the Conclusion says nesting material reduces the effects of thermal stress. It does not establish that bedding material itself affects behavioral responses to stress, or extend the finding beyond this thermal-stress context. The separate pain claim is attributed to [185], not this source. (LLM, conf=0.84)

  cites [184] BN Gaskill, 2017, Stressed out: providing laboratory animals with behavioral control to reduce the physiological effects of stress, Lab Animal sourceExperimental — accuracy not yet measured

In the paper

“Cage/tank/housing system (type and dimensions) | Affects behaviour [177] and fear learning [178].”

What the sources say

• Partly supported

  The Fear Learning section (p. 15) reports different fear-potentiated startle performance across housing types in male B6J mice, supporting an effect of housing system on this measure of fear learning. The Abstract (p. 2) says the cages compared were the same size, so the paper does not establish an effect of cage dimensions. (LLM, conf=0.86)

  cites [178] M Kallnik, 2007, Impact of IVC housing on emotionality and fear learning in male C3HeB/FeJ and C57BL/6J mice, Mamm Genome sourceExperimental — accuracy not yet measured

In the paper

“In systematic reviews, initial screens to identify papers are based on titles, abstracts, and keywords [156].”

What the sources say

• Partly supported

  The Methods section (Step 1; page unavailable) describes an initial screening decision made with each record’s title and abstract displayed. It does not say that keywords were used in that screening, so the three-part claim is only partly supported. (LLM, conf=0.87)

  cites [156] A Bannach-Brown, 2019, Machine learning algorithms for systematic review: reducing workload in a preclinical review of animal studies and reducing human screening error, Syst Rev sourceExperimental — accuracy not yet measured

In the paper

“Underreporting of welfare management procedures contributes to the perpetuation of noncompliant methodologies and insufficient or inappropriate use of analgesia [213] or other welfare measures.”

What the sources say

• Partly supported

  The Conclusion says that under-reporting pain management encourages under-treatment and can lead researchers to repeat apparent analgesic non-use. The Introduction likewise describes how that impression can perpetuate non-use. The paper does not establish that the resulting methods are noncompliant or extend this finding to other welfare measures. (LLM, conf=0.85)

  cites [213] L Carbone, 2016, Pain and laboratory animals: publication practices for better data reproducibility and better animal welfare, PLoS ONE sourceExperimental — accuracy not yet measured

In the paper

“Adds an additional source of variation [189,190].”

What the sources say

• Supported

  In the main text (page unavailable), the paper reports that cage-cleaning frequency affects rats’ behavioral and biological stress responses and recommends keeping it consistent across treatment groups. This supports treating sanitation practices as a source of experimental variation. (LLM, conf=0.90)

  cites [189] E Prager, 2011, The importance of reporting housing and husbandry in rat research, Front Behav Neurosci source
• ?Not checked — no full text of the sourcecites [190] MD Rosenbaum, 2009, Effects of cage-change frequency and bedding volume on mice and their microenvironment, Journal of the American Association for Laboratory Animal Science: JAALAS

In the paper

“Social environment (group size and composition/stocking density) | Compromises animal welfare [191]. Induces aggressive behaviour [192,193] and stress [180].”

What the sources say

• Partly supported

  Section 4.3 reports that aggression increased substantially with group size in male mice. Sections 4.2 and 5 also find that cage size affected aggression, but smaller cages reduced it in this study. The paper does not test group composition or support a general claim that higher stocking density increases aggression. (LLM, conf=0.86)

  cites [192] PLP Van Loo, 2001, Modulation of aggression in male mice: influence of group size and cage size, Physiol Behav sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [193] CE Adams, 2007, Multiple determinants of welfare in farmed fish: stocking density, disturbance, and aggression in Atlantic salmon (Salmo salar), Canadian Journal of Fisheries and Aquatic Sciences source

In the paper

“Environmental enrichment | Reduces anxiety [196,197], stress [196,197], and abnormal repetitive behaviour [198–200]. Reduces susceptibility to epilepsy [201] and osteoarthritis [202] and modifies the pathology of neurological disorders [203]. Increases foraging behaviour in fish [204].”

What the sources say

• Supported

  The Replicability section says successful enrichments often reduce or prevent abnormal behaviors. The Mouse Enrichment: Brief Commentary section gives examples in which enrichment reduced stereotypy in mice and bank voles. The paper cautions that enrichment may have less effect on some stereotypies once they have developed. (LLM, conf=0.92)

  cites [198] JP Garner, 2005, Stereotypies and other abnormal repetitive behaviors: potential impact on validity, reliability, and replicability of scientific outcomes, ILAR J source
• ?Not checked — no full text of the sourcecites [199] AN-M Gross, 2011, Simply a nest? Effects of different enrichments on stereotypic and anxiety-related behaviour in mice, Appl Anim Behav Sci source
• ?Not checked — no full text of the sourcecites [200] H Wurbel, 2001, Ideal homes? Housing effects on rodent brain and behaviour, Trends Neurosci source

In the paper

“Social environment (group size and composition/stocking density) | Compromises animal welfare [191]. Induces aggressive behaviour [192,193] and stress [180].”

What the sources say

• Partly supported

  Section 4.4 links group size and male-mouse social relationships to aggression and welfare problems. But the Conclusions say housing practices can either compromise or enhance welfare, depending on the situation. The categorical claim extends beyond that context-dependent finding, and the cited passages do not establish that stocking density itself compromises welfare. (LLM, conf=0.88)

  cites [191] S Kappel, 2017, To group or not to group? Good practice for housing male laboratory mice, Animals sourceExperimental — accuracy not yet measured

In the paper

“Untreated pain can affect the animals’ biology and add variability to the experiment; however, specific pain management procedures can also introduce variability, affecting experimental data [212,213].”

What the sources say

• Supported

  The Focus on Reproducibility Review states that the chosen pain-management protocol can affect scientific results and increase data variability. Its Effect of analgesia protocols section also explains how gaps in analgesia can contribute to variability. (LLM, conf=0.94)

  cites [212] P Jirkof, 2017, Side effects of pain and analgesia in animal experimentation, Lab Anim (NY) source
• Supported

  The Introduction states that pain-relieving drugs can skew experimental data by introducing unwanted variability. The Discussion section on pain management as a methodological concern adds that research outcomes can differ depending on which analgesic is used. (LLM, conf=0.94)

  cites [213] L Carbone, 2016, Pain and laboratory animals: publication practices for better data reproducibility and better animal welfare, PLoS ONE source

In the paper

“Access to raw data also facilitates text and automated data mining [246].”

What the sources say

• Partly supported

  The Abstract (Methods) describes mining mouse-behavior data for predictive patterns, supporting the automated data-mining part of the claim. Neither that passage nor the Introduction and Discussion passages addresses access to raw data or text mining, so the broader claim is not fully supported. (LLM, conf=0.79)

  cites [246] N Kafkafi, 2014, Mining mouse behavior for patterns predicting psychiatric drug classification, Psychopharmacology (Berl) sourceExperimental — accuracy not yet measured

In the paper

“For example, bias, as indicated by a prevalence of more favourable outcomes, has been demonstrated for clinical research funded by industry compared with studies funded by other sources [257–259].”

What the sources say

• Supported

  The Results report that industry-sponsored studies more often had favourable conclusions than non-industry-sponsored studies (RR 1.34, 95% CI 1.19–1.51). The discussion (pp. 18–19) also describes more favourable results and conclusions as an industry-associated bias. (LLM, conf=0.93)

  cites [257] A Lundh, 2012, Industry sponsorship and research outcome, Cochrane Database Syst Rev source
• Supported

  The Results report that industry-associated dental implant trials had significantly lower reported annual failure rates than non-industry trials. The Discussion describes this as evidence that funding source affects reported outcomes, while the Introduction notes similar associations in pharmaceutical clinical trials. (LLM, conf=0.91)

  cites [258] A Popelut, 2010, Relationship between sponsorship and failure rate of dental implants: a systematic approach, PLoS ONE source
• Supported

  The Discussion (page not provided) says drug-industry-sponsored research was more likely to produce results favouring the sponsor’s product than research funded by other sources. It also describes the pattern as evidence of systematic bias in published research. (LLM, conf=0.94)

  cites [259] J Lexchin, 2003, Pharmaceutical industry sponsorship and research outcome and quality: systematic review, BMJ source

In the paper

“Registering a protocol before the start of the experiment enables researchers to demonstrate that the hypothesis, approach, and analysis were planned in advance and not shaped by data as they emerged; it enhances scientific rigour and protects the researcher against concerns about selective reporting of results [239,240].”

What the sources say

• Supported

  The Discussion’s Implications section says trial registration records the study’s purpose, design, and prespecified outcomes. The Discussion also says preregistration essentially eliminated opportunities to change a hypothesis’s direction, significance threshold, or composite variables after the fact; the Results explain how declaring an outcome in advance constrains selective reporting. The paper discusses these safeguards in the context of clinical trials. (LLM, conf=0.86)

  cites [239] RM Kaplan, 2015, Likelihood of null effects of large NHLBI clinical trials has increased over time, PLoS ONE source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [240] C Allen, 2018, Open Science challenges, benefits and tips in early career and beyond, PsyArXiv

In the paper

“Akin to the approach taken for clinical trials, protocol registration has emerged as a mechanism that is likely to improve the transparency of animal research [232,237,238].”

What the sources say

• ?Not checked — no full text of the sourcecites [232] I Chalmers, 2014, How to increase value and reduce waste when research priorities are set, The Lancet source
• Partly supported

  The Conclusions (page unavailable) describe preregistration as growing in basic science and as an incentive for transparent practices. The discussion of clinical trial registration (page unavailable) also supports the analogy to clinical trials. The paper does not specifically establish that protocol registration improves transparency in animal research. (LLM, conf=0.83)

  cites [237] CD Chambers, 2014, Instead of "playing the game" it is time to change the rules: Registered Reports at AIMS Neuroscience and beyond, AIMS Neuroscience sourceExperimental — accuracy not yet measured
• Partly supported

  The Closing (p. 4) describes preregistration as part of Registered Reports and links it to transparency practices. Open Science Practices in This Special Issue (p. 3) describes registered proposals and materials as practices that make research conduct explicit. The paper discusses social psychology research, however, and does not establish that protocol registration is likely to improve transparency specifically in animal research. (LLM, conf=0.85)

  cites [238] BA Nosek, 2014, Registered Reports a method to increase the credibility of published results, Soc Psychol sourceExperimental — accuracy not yet measured

In the paper

“Journal articles with accompanying data may be cited more frequently [248,249].”

What the sources say

• Supported

  The Results report that publicly available trial data was associated with a 69% higher citation count. The Discussion specifies that the study examined cancer clinical trials sharing microarray data and cautions that the association does not establish causation. The claim's 'may' is consistent with those findings. (LLM, conf=0.92)

  cites [248] HA Piwowar, 2007, Sharing detailed research data is associated with increased citation rate, PLoS ONE source
• Supported

  The section “Gain more citations and visibility by sharing data” reports citation advantages for journal articles whose data were shared or linked. That supports the claim’s cautious wording, “may be cited more frequently”; the paper reports an association, not a guaranteed or causal effect. The supplied passage has no page number. (LLM, conf=0.94)

  cites [249] EC McKiernan, 2016, How open science helps researchers succeed, Elife source

In the paper

“Evidence for preclinical research also indicates that funding sources may influence the interpretation of study outcomes [254,260].”

What the sources say

• Supported

  The Conclusions (page unavailable) report that industry-sponsored animal studies were less likely than non-industry-sponsored studies to conclude that atrazine was harmful. The authors summarize that industry-sponsored studies were more likely to report findings favoring the sponsor, supporting the hedged claim that funding sources may influence how preclinical outcomes are interpreted. (LLM, conf=0.90)

  cites [254] L Bero, 2016, The relationship between study sponsorship, risks of bias, and research outcomes in atrazine exposure studies conducted in non-human animals: Systematic review and meta-analysis, Environ Int source
• Partly supported

  The Discussion reports that nonindustry-sponsored preclinical statin studies yielded greater efficacy estimates than industry-sponsored studies, and the Conclusion identifies funding source as an influence on preclinical research. This supports an association between funding and reported outcomes, but the paper does not directly examine how researchers interpret those outcomes. (LLM, conf=0.82)

  cites [260] D Krauth, 2014, Nonindustry-Sponsored Preclinical Studies on Statins Yield Greater Efficacy Estimates Than Industry-Sponsored Studies: A Meta-Analysis, PLoS Biol sourceExperimental — accuracy not yet measured

In the paper

“There are several types of studies in which a priori sample size calculations are not appropriate.”

In the paper

“When the experiment tests the effect of an intervention on the mean of a continuous outcome measure, the sample size can be calculated a priori, based on a mathematical relationship between the predefined, biologically relevant effect size, variability estimated from prior data, chosen significance level, power, and sample size (see Box 3 and [17,41] for practical advice).”

What the sources say

• ?Not checked — no full text of the sourcecites [17] ST Bate, 2014 source
• ?Not checked — no full text of the sourcecites [41] MFW Festing, 2018, On determining sample size in experiments involving laboratory animals, Lab Anim source

In the paper

“Systematic reviews and previous studies are more appropriate sources of information on variability [45].”

What the sources say

• Not checked yetcites [45] Bate ST. How to decide your sample size when the power calculation is not straightforward. 2018 Aug 1 [cited 2018 Aug 2]. In: NC3Rs.org.uk [Internet]. Available from: https://www.nc3rs.org.uk/news/how-decide-your-sample-size-when-power-calculation-not-straightforward.

In the paper

“Sample sizes that are too small (i.e., underpowered studies) produce inconclusive results, whereas sample sizes that are too large (i.e., overpowered studies) raise ethical issues over unnecessary use of animals and may produce trivial findings that are statistically significant but not biologically relevant [35].”

What the sources say

• ?Not checked — no full text of the sourcecites [35] MD Jennions, 2003, A survey of the statistical power of research in behavioral ecology and animal behavior, Behav Ecol source

In the paper

“Low power has three effects: first, within the experiment, real effects are more likely to be missed; second, when an effect is detected, this will often be an overestimation of the true effect size [24]; and finally, when low power is combined with publication bias, there is an increase in the false positive rate in the published literature [36].”

In the paper

“Following the use of a one-sided test, the investigator cannot then test for the possibility of missing an effect in the untested direction.”

In the paper

“Pilot studies alone are unlikely to provide adequate data on variability for a power calculation for future experiments.”

What the sources say

• Not checked yetcites [45] Bate ST. How to decide your sample size when the power calculation is not straightforward. 2018 Aug 1 [cited 2018 Aug 2]. In: NC3Rs.org.uk [Internet]. Available from: https://www.nc3rs.org.uk/news/how-decide-your-sample-size-when-power-calculation-not-straightforward.

In the paper

“Reporting the precise method used to allocate animals or experimental units to groups enables readers to assess the reliability of the results and identify potential limitations.”

In the paper

“Inappropriate data cleaning has the potential to bias study outcomes [53]; providing the reasoning for removing data points enables the distinction to be made between responsible data cleaning and data manipulation.”

What the sources say

• ?Not checked — no full text of the sourcecites [53] S Scott, 2008, Design, power, and interpretation of studies in the standard murine model of ALS, Amyotroph Lateral Scler source

In the paper

“It can introduce bias that influences the results, as a researcher may (consciously or subconsciously) make judgements in allocating an animal to a particular group, or because of unknown and uncontrolled differences in the experimental conditions or animals in different groups.”

In the paper

“Researchers often expect a particular outcome and can unintentionally influence the experiment or interpret the data in such a way as to support their preferred hypothesis [81].”

What the sources say

• ?Not checked — no full text of the sourcecites [81] R Nuzzo, 2015, How scientists fool themselves–and how they can stop, Nature News source

In the paper

“In studies in which sample sizes are small, simple randomisation may result in unbalanced groups”

In the paper

“This can be avoided by using allocation concealment (see Item 5. Blinding).”

In the paper

“Although there is primary evidence of the impact of blinding in the clinical literature that directly compares blinded versus unblinded assessment of outcomes [82], there is limited empirical evidence in animal research [83,84].”

What the sources say

• ?Not checked — no full text of the sourcecites [83] R Rosenthal, 1963, The effect of experimenter bias on the performance of the albino rat, Behav Sci source
• ?Not checked — no full text of the sourcecites [84] R Rosenthal, 1964, A longitudinal study of the effects of experimenter bias on the operant learning of laboratory rats, J Psychiatr Res source

In the paper

“It is especially important to use randomisation in situations in which it is not possible to blind all or parts of the experiment, but even with randomisation, researcher bias can pervert the allocation.”

In the paper

“Thus, the use of randomisation is a prerequisite for any experiment designed to test a hypothesis.”

In the paper

“Selecting an animal ‘at random’ (i.e., haphazardly or arbitrarily) from a cage is not statistically random, as the process involves human judgement.”

In the paper

“In hypothesis-testing studies using inferential statistics, investigators frequently confuse statistical significance and small p-values with biological or clinical importance [149].”

What the sources say

• ?Not checked — no full text of the sourcecites [149] RL Wasserstein, 2019, Moving to a World Beyond “p < 0.05”, The American Statistician source

In the paper

“Physiological acclimatisation after a stressful event, such as transport (e.g., between supplier, animal facility, operating theatre, and laboratory), but before the experiment begins allows stabilisation of physiological responses of the animal [137,138].”

What the sources say

• ?Not checked — no full text of the sourcecites [137] AM Holmes, 2018, Effects of transportation, transport medium and re-housing on Xenopus laevis (Daudin), Gen Comp Endocrinol source
• ?Not checked — no full text of the sourcecites [138] LA Conour, 2006, Preparation of animals for research—issues to consider for rodents and rabbits, ILAR J source

In the paper

“However, statistical significance is heavily influenced by sample size and variation in the data (see Item 2. Sample size).”

In the paper

“Reporting the characteristics of all animals used is equivalent to standardised human patient demographic data; these data support both the internal and external validity of the study results.”

In the paper

“Food and water (type, composition, supplier, and access) | Affects body weight, tumour development, nephropathy severity [181], and the threshold for developing parkinsonian symptoms [182].”

What the sources say

• ?Not checked — no full text of the sourcecites [181] JK Haseman, 2003, Effect of diet and animal care/housing protocols on body weight, survival, tumor incidences, and nephropathy severity of F344 rats in chronic studies, Toxicol Pathol source

In the paper

“Leaving out of the abstract information such as the species of animal used or the drugs being tested limits the value of preclinical systematic reviews as relevant studies cannot be identified and included.”

In the paper

“When multiple independent studies included in a meta-analysis show quantitatively similar effects, even if each is statistically nonsignificant, this provides powerful evidence that a relationship is ‘real’, although small.”

In the paper

“Food and water (type, composition, supplier, and access) | Affects body weight, tumour development, nephropathy severity [181], and the threshold for developing parkinsonian symptoms [182].”

What the sources say

• ?Not checked — no full text of the sourcecites [181] JK Haseman, 2003, Effect of diet and animal care/housing protocols on body weight, survival, tumor incidences, and nephropathy severity of F344 rats in chronic studies, Toxicol Pathol source

In the paper

“Food and water (type, composition, supplier, and access) | Affects body weight, tumour development, nephropathy severity [181], and the threshold for developing parkinsonian symptoms [182].”

What the sources say

• ?Not checked — no full text of the sourcecites [181] JK Haseman, 2003, Effect of diet and animal care/housing protocols on body weight, survival, tumor incidences, and nephropathy severity of F344 rats in chronic studies, Toxicol Pathol source

In the paper

“Bedding and nesting material | Affects behavioural responses to stress [184] and pain [185].”

What the sources say

• ?Not checked — no full text of the sourcecites [185] I Robinson, 2004, Development of neuropathic pain is affected by bedding texture in two models of peripheral nerve injury in rats, Neurosci Lett source

In the paper

“Thus, results from well-designed hypothesis-testing studies provide stronger evidence than those from exploratory or descriptive studies.”

• Also stated in: § Item 18. Generalisability/translation · plain

  “Well-designed hypothesis-testing studies provide more robust evidence than exploratory studies (see Item 13. Objectives).”

In the paper

“Sanitation (frequency of cage/tank water changes, material transferred, water quality) | Affects blood pressure, heart rate, behaviour [188].”

What the sources say

• ?Not checked — no full text of the sourcecites [188] JL Duke, 2001, The effects of routine cage-changing on cardiovascular and behavioral parameters in male Sprague-Dawley rats, Contemp Top Lab Anim Sci

In the paper

“Reporting effect sizes with confidence intervals also facilitates extraction of useful data for systematic review and meta-analysis.”

In the paper

“Social environment (group size and composition/stocking density) | Compromises animal welfare [191]. Induces aggressive behaviour [192,193] and stress [180].”

What the sources say

• ?Not checked — no full text of the sourcecites [180] M Pavlidis, 2013, Husbandry of zebrafish, Danio rerio, and the cortisol stress response, Zebrafish source

In the paper

“Environmental enrichment | Reduces anxiety [196,197], stress [196,197], and abnormal repetitive behaviour [198–200]. Reduces susceptibility to epilepsy [201] and osteoarthritis [202] and modifies the pathology of neurological disorders [203]. Increases foraging behaviour in fish [204].”

What the sources say

• ?Not checked — no full text of the sourcecites [203] AJ Hannan, 2014, Environmental enrichment and brain repair: harnessing the therapeutic effects of cognitive stimulation and physical activity to enhance experience-dependent plasticity, Neuropathol Appl Neurobiol source

In the paper

“Environmental enrichment | Reduces anxiety [196,197], stress [196,197], and abnormal repetitive behaviour [198–200]. Reduces susceptibility to epilepsy [201] and osteoarthritis [202] and modifies the pathology of neurological disorders [203]. Increases foraging behaviour in fish [204].”

What the sources say

• ?Not checked — no full text of the sourcecites [196] P Chapillon, 1999, Rearing environmental enrichment in two inbred strains of mice: 1. Effects on emotional reactivity, Behav Genet source
• ?Not checked — no full text of the sourcecites [197] TR Hendershott, 2016, Effects of environmental enrichment on anxiety-like behavior, sociability, sensory gating, and spatial learning in male and female C57BL/6J mice, Behav Brain Res source

In the paper

“Sex of the experimenter | Affects physiological stress and pain behaviour [205].”

What the sources say

• ?Not checked — no full text of the sourcecites [205] RE Sorge, 2014, Olfactory exposure to males, including men, causes stress and related analgesia in rodents, Nature Methods source

In the paper

“Lighting (type, schedule, and intensity) | Modifies immune and stress responses [195].”

What the sources say

• ?Not checked — no full text of the sourcecites [195] RT Dauchy, 2011, Eliminating animal facility light-at-night contamination and its effect on circadian regulation of rodent physiology, tumor growth, and metabolism: a challenge in the relocation of a cancer research laboratory, J Am Assoc Lab Anim Sci

In the paper

“Sanitation (frequency of cage/tank water changes, material transferred, water quality) | Affects blood pressure, heart rate, behaviour [188].”

What the sources say

• ?Not checked — no full text of the sourcecites [188] JL Duke, 2001, The effects of routine cage-changing on cardiovascular and behavioral parameters in male Sprague-Dawley rats, Contemp Top Lab Anim Sci

In the paper

“Sanitation (frequency of cage/tank water changes, material transferred, water quality) | Affects blood pressure, heart rate, behaviour [188].”

What the sources say

• ?Not checked — no full text of the sourcecites [188] JL Duke, 2001, The effects of routine cage-changing on cardiovascular and behavioral parameters in male Sprague-Dawley rats, Contemp Top Lab Anim Sci

In the paper

“Environmental enrichment | Reduces anxiety [196,197], stress [196,197], and abnormal repetitive behaviour [198–200]. Reduces susceptibility to epilepsy [201] and osteoarthritis [202] and modifies the pathology of neurological disorders [203]. Increases foraging behaviour in fish [204].”

What the sources say

• ?Not checked — no full text of the sourcecites [196] P Chapillon, 1999, Rearing environmental enrichment in two inbred strains of mice: 1. Effects on emotional reactivity, Behav Genet source
• ?Not checked — no full text of the sourcecites [197] TR Hendershott, 2016, Effects of environmental enrichment on anxiety-like behavior, sociability, sensory gating, and spatial learning in male and female C57BL/6J mice, Behav Brain Res source

In the paper

“Lighting (type, schedule, and intensity) | Modifies immune and stress responses [195].”

What the sources say

• ?Not checked — no full text of the sourcecites [195] RT Dauchy, 2011, Eliminating animal facility light-at-night contamination and its effect on circadian regulation of rodent physiology, tumor growth, and metabolism: a challenge in the relocation of a cancer research laboratory, J Am Assoc Lab Anim Sci

In the paper

“Sex of the experimenter | Affects physiological stress and pain behaviour [205].”

What the sources say

• ?Not checked — no full text of the sourcecites [205] RE Sorge, 2014, Olfactory exposure to males, including men, causes stress and related analgesia in rodents, Nature Methods source

In the paper

“Biosecurity (level) | The microbiological status of animals causes variation in systemic disease parameters [194].”

What the sources say

• ?Not checked — no full text of the sourcecites [194] A Bleich, 2012, Time to include the gut microbiota in the hygienic standardisation of laboratory rodents, Comp Immunol Microbiol Infect Dis source