Shownotes

Joe’s background

03:03 -

• I consider myself first and foremost a runner. I began my athletic career as a track athlete, but in recent years I have transitioned to ultra-trail running as a way to explore something new. Professionally, I have been a qualified sports scientist for approximately 15 years and currently teach at Technological University Dublin.
• Around five or six years ago, I launched the Sports Science Replication Center, an initiative focused on meta-science research within the field of sports science.
• My aim with this project has been to examine and improve research methodologies in the discipline, drawing on lessons from other fields—particularly psychology—where, over the past decade, significant concerns have been raised about the way research is conducted.
• These issues within the research ecosystem inspired me to address similar concerns in sports and exercise science. While the concept was new to our field, it was heavily influenced by the replication initiatives already established elsewhere.
• My own research background was originally in footwear, specifically examining the performance and injury implications of running shoes, which was the subject of my PhD. However, in recent years, I have shifted my focus entirely toward research methods and the broader scientific approach within sports science.
• On a lighter note, my favorite road running shoe of all time was the Saucony Hattori. It was only in production for about a year, but it stood out to me for being incredibly light, soft, and comfortable—almost like wearing a slipper.
• As for trail running, I am still exploring what works best for me, but I have a strong preference for one of the original Inov-8 models. Its high technicality, thin and flexible construction, and excellent ground feel make it ideal for fast, technical descents.

Concerns about Sports science research

05:30 -

• The concerns I have been addressing in my work stem from a broader philosophy of science, one that dates back centuries.
• Francis Bacon once wrote that scientists should not be motivated by fun, feelings of superiority, fame, power, or money, but instead should pursue science to make the world a better place. This was expressed in the 1600s, yet it remains a guiding ideal today.
• At its core, science should be a rigorous, objective process aimed at understanding the world, uncovering truth, and developing knowledge without bias. However, this process has become distorted by the way research now functions as a form of currency.
• Journals, for example, are incentivized to publish work that will attract readers, citations, and ultimately subscriptions from universities and institutions. The more novel or sensational the research, the more attention it garners—similar to the concept of clickbait—which in turn boosts a journal’s impact and revenue.
• Researchers are subject to similar pressures. In academia, the number of publications and citation counts often serve as benchmarks for securing tenure, promotions, or research funding.
• As a result, there is a strong incentive to publish as frequently as possible. I have noticed that in my career, I have been asked over a hundred times how many papers I have published, but rarely—if ever—what the most meaningful piece of research I have conducted has been. This “publish or perish” culture encourages quantity over quality.
• When journals prioritize novelty and researchers are motivated to produce work rapidly, the result is a research ecosystem that values exciting findings over rigorous, reproducible science. This imbalance has led to what is now referred to as the “replication crisis.”
• This term originated in social psychology, where large-scale efforts to replicate existing studies revealed alarmingly low replication rates. Replication, a fundamental cornerstone of science, is the process of repeating a study’s methods to see if similar results can be obtained. If a claim is sound, others should be able to reproduce it reliably.
• The low replication rates observed suggested that the current ecosystem is, in some ways, undermining science. Instead of consistently advancing knowledge, questionable research practices and systemic pressures are leading us away from scientific quality and toward inflated publication counts.
• This has significant implications for both the short- and long-term credibility of science, and it is a key issue that I believe needs to be addressed before we can expect lasting trust and confidence in research.

Replication crisis

09:35 -

• Methodologically, several underlying issues contribute to the replication crisis, and one of the most prominent is publication bias. This bias means that studies reporting novel, statistically significant, or groundbreaking findings are far more likely to be published than those producing non-significant or confirmatory results. As a consequence, researchers are incentivized to focus on uncovering “the next new thing” rather than slowing down to validate and reaffirm existing knowledge.
• This pursuit of novelty is reinforced by the publication infrastructure itself. Journals prefer to publish results that are exciting or unexpected because they attract more readers and citations, which, in turn, increase the journal’s standing and revenue. Researchers, operating in a system where publications are the currency for career advancement, naturally align their work with these incentives.
• It is important to note that I am not suggesting fraudulent activity is widespread. In my view, deliberate data fabrication or gross manipulation remains rare.
• The more common issue is that the research ecosystem has normalized certain questionable practices simply because they have been in use for so long. One of the clearest examples is the widespread acceptance of small sample sizes.
• In sports and exercise science, this problem is especially acute. Much of the research involves human participants, which requires significant time, resources, and effort. Recruiting participants can be difficult, and dropouts are common.
• These challenges have led to a norm where studies rely on small, convenient samples, even when investigating broad or ambitious claims. The problem is that small samples produce data that is more variable and less reliable, making the resulting claims harder to reproduce with confidence.
• Compounding this issue is the tendency—whether conscious or subconscious—for researchers to make decisions that increase the likelihood of finding statistically significant results. This is often driven by human bias toward novelty.
• For example, when teaching undergraduates, I frequently remind them that the value of their study lies not in confirming a particular outcome, but in uncovering the truth—whether or not it is statistically significant. In a hypothetical study on whether shaving body hair improves swimming performance in triathletes, the aim should not be to prove a benefit, but rather to determine the actual effect, if any.
• Unfortunately, because significant, novel results are more likely to be published, the literature becomes skewed. This distorts the evidence base and makes it harder for practitioners to know what actions to take.
• In the case of the shaving example, it may be unclear whether an intervention is truly effective or whether its apparent benefit is simply the result of biased research practices. This erosion of clarity is one of the key reasons confidence in scientific findings is undermined.

Meaning of a significant finding versus a non-significant finding

14:02 -

• When discussing statistical significance, it is important to clarify exactly what it means—and equally important, what it does not mean—particularly in the context of small sample sizes and low statistical power.
• In science, a study begins with a hypothesis or working theory, which is then tested as rigorously as possible while controlling for confounding factors.
• For example, in a hypothetical study examining whether shaving the arms and legs improves swimming performance, we would control for environmental conditions such as water temperature, testing format, and so forth. Data collected from participants would then undergo statistical inference, producing a probability measure known as the p-value.
• The p-value indicates the likelihood that the observed results could be replicated in future trials, assuming similar conditions. A common threshold for “statistical significance” is p < 0.05, which suggests that the probability of the result occurring purely by chance is low.
• However, this is not a statement of fact—it is a probabilistic estimate. Unfortunately, both researchers and especially research consumers often misinterpret statistical significance as proof that a finding is definitively true, which is not the case.
• The level of confidence in a finding depends heavily on statistical power, which is a function of both the sample size and the size of the effect being measured.
• With small samples—such as testing five swimmers with shaved limbs and five without—the variability in the data is high, and confidence in the result is low. Larger effects, like the performance difference between a mountain bike and a time trial bike, require smaller sample sizes to detect reliably, while subtle effects demand much larger participant numbers to ensure confidence.
• In reality, large samples are often impractical in sports science due to resource limitations and participant availability. This is why replication is so vital.
• No single study proves anything; confidence grows only when multiple independent studies produce consistent results in the same direction. This is also why meta-analyses, which aggregate data from multiple studies, can provide a clearer picture of the true effect.
• However, there is a major complication: publication bias. Studies with statistically significant or novel findings are far more likely to be published than those with non-significant results, even though the latter are equally important for understanding reality.
• This means that the published literature often represents a skewed subset of all research conducted. The “file drawer effect” describes how many studies—often more than half—remain unpublished because they found no significant effect or did not produce an outcome deemed interesting enough for journals.
• This bias distorts the evidence base, making it difficult for practitioners, athletes, or coaches to form truly evidence-based decisions. Even if someone reviews all available literature on a topic, they may still be seeing only a partial, selective picture of the actual research that has been done.

Reasons for not doing some research

22:22 -

• There are multiple reasons why non-significant findings often go unpublished, and they operate at several levels.
• From the researcher’s perspective, human bias plays a large role. Many researchers, even if they know better intellectually, tend to feel more satisfaction in reporting a statistically significant result.
• There is also a practical consideration: experience teaches us that studies without significant findings are far less likely to be accepted by journals, making the time and effort of writing, submission, and revision feel like a poor investment.
• Beyond the researcher’s own motivation, systemic barriers reinforce this trend. Peer reviewers may undervalue studies that fail to support existing theories or produce significant results, sometimes even stating outright that the work would have been more publishable if a significant effect had been found.
• Editors, too, operate under the pressures of the “clickbait” economy in academic publishing. Journals are more likely to attract readership and citations with novel, statistically significant findings, so editors tend to prioritize those submissions—although it is worth noting that many journals have been making efforts to improve in this regard.
• The end result is that, from initial submission to final editorial decision, there are multiple points where non-significant results face lower odds of publication.
• This is not because such findings lack scientific value—they can be equally important for building a reliable evidence base—but because the current publishing ecosystem still favours results that are statistically significant, novel, and more likely to capture attention.

Comments on published papers with only statistically significant findings

25:03 -

• I became aware of these issues through a combination of statistical reasoning, literature analysis, and my own experience reviewing and conducting research. One of the most striking indicators was the observation—published in one of my earlier papers—that around 80% of findings in sports science are statistically significant.
• This is highly suspicious when you consider two facts: most effects studied in our field are likely to be small or medium in size, and most studies use small sample sizes.
• From a purely statistical perspective, small samples have low power, meaning that even if a true significant effect exists, the probability of detecting it is relatively low. In theory, this should result in a large proportion of non-significant results.
• Yet the literature shows the opposite—an unusually high proportion of significant findings—which suggests strong publication bias or selective reporting. It’s an “iceberg” situation: the visible, published studies are only the tip, while a much larger body of non-significant findings remains unseen.
• This mismatch between what the mathematics predicts and what the literature reports was one of the earliest red flags for me. Even before we conducted our large-scale replication work—which has since confirmed the scope of the problem—I had publicly predicted a 75% failure rate in replication attempts within sports science.
• That estimate came directly from recognising the statistical impossibility of so many significant results given our field’s methodological norms.
• To be clear, I do not believe this is the result of widespread intentional misconduct. Rather, these patterns reflect a research culture that has, over time, normalised practices that favour the publication of significant, novel results over robust, reproducible science.
• There are, however, encouraging signs of change. Some journals, such as the Journal of Sports Sciences, now require mandatory sample size justification for submitted studies.
• The Journal of Science and Medicine in Football has adopted similar measures, and others have implemented pre-registration requirements, where researchers must submit their study design before collecting data.
• In my recent experience as a peer reviewer for Sports Medicine, I was explicitly instructed to review the raw data and assess the study’s reproducibility and replication potential—an important step toward improving research quality. These reforms give me hope that the field is moving in a better direction.

Success and failure rate of the replications

27:39 -

• Recently, my colleagues and I published a large-scale replication study in Sports Medicine that took more than four and a half years to complete. It was the first systematic attempt to randomly select and replicate as much research in sports and exercise science as possible.
• The project involved over 100 collaborators, making it one of the largest undertakings of its kind in our field.
• Our findings were sobering: only 28% of the studies we attempted to replicate produced results consistent with the originals, meaning there was a 72% failure rate.
• It is important to stress that a failed replication does not automatically mean the original finding was false. Replication is itself subject to error, and outcomes can differ for many reasons. However, low replication rates do mean we should be cautious about the level of confidence we place in original findings.
• One of the most important—and perhaps most concerning—patterns we observed was a substantial reduction in reported effect sizes. On average, when we replicated studies with larger sample sizes, the effect was 75% smaller than originally claimed.
• For practitioners, coaches, and athletes, this matters greatly. If an initial study reports that caffeine improves a 20 km cycling time trial by 4%, the real-world effect—based on a more robust replication—may be closer to 1%.
• This inflation of effect sizes is partly due to the high variability present in small-sample research. A single, small study can produce extreme results simply by chance, and these results are more likely to be published.
• Over time, such inflated claims can become entrenched in coaching practices and athlete decision-making, even when the true effect is far smaller. A famous example is the Specialized wind tunnel test on leg shaving for cyclists, which reported a time saving of roughly 60–90 seconds over 40 km for a single professional triathlete.
• This figure has been widely cited, but with a larger, more representative sample, the true benefit would almost certainly be smaller.
• In some cases—such as leg shaving—the consequences of inflated claims are relatively harmless. However, in other contexts, such as supplement use or medical treatments, overstated effects can have serious health implications.
• This issue is not unique to sports science; similar replication and effect-size inflation problems have been documented in cancer biology, economics, psychology, and pharmacology. One notable example involved an antidepressant that was initially reported to be highly effective, but replication studies failed to confirm this.
• Despite this, the drug was sold over the counter, leaving patients with serious mental health conditions using an ineffective treatment.
• In sports science, while some claims may appear trivial, the underlying principle is not.
• Sport has a profound impact on people’s lives, and as scientists, we have a professional responsibility to ensure that the claims we make—knowing they will be interpreted as truth by coaches and athletes—are robust, reproducible, and based on the best available evidence.

Practical applications of these issues

34:09 -

• For athletes, coaches, and others who want to use sports science to inform their practice, my first recommendation is to focus on the fundamentals that are unquestionably important for performance.
• Before worrying about marginal gains, make sure the core principles are firmly in place—adequate and progressive training load, proper recovery, sound nutrition, good sleep habits, and specificity in training. These factors alone will take you most of the way toward your potential, provided you also have the genetic capacity to adapt.
• Too often, people get distracted by the promise of a shortcut—whether it’s a supplement, a piece of equipment, or a niche training method—hoping it will transform them from middle-of-the-pack to elite.
• The reality is that these “fluffy” elements contribute far less to performance than people think, especially if the basics are not yet fully optimised.
• I agree that not every traditional training concept is equally well-supported by high-quality controlled research. Periodisation, for example, lacks a universally proven model in the literature, particularly when comparing different sports and competitive structures.
• My point is less about a specific template and more about the underlying principle: the process of applying a training stimulus, allowing recovery, and progressively increasing load over time.
• This pattern—whether you call it periodisation or not—is essential for adaptation. In practice, the best periodisation strategy is often highly individualised, guided by how an athlete is responding week to week.
• There are, however, areas where the research is both strong and consistent. One clear example is high-intensity interval training (HIIT). Across numerous studies, HIIT has been shown to improve VO₂max, threshold power, time trial performance, and other key endurance metrics.
• While the optimal structure can vary between individuals, the general principle that HIIT drives meaningful physiological improvements is well established.
• In short, the best way to consume sports science as a practitioner is to: 
  • Anchor your approach in the evidence-backed fundamentals.
  • Recognise where science provides strong consensus versus where it is less conclusive.
  • Individualise application based on the athlete’s specific context, needs, and responses.

How to evaluate a specific topic in scientific literature

38:51 -

• When assessing a specific topic—such as carbohydrate consumption and endurance performance—the challenge is that even with expertise, it can be very difficult to know how much trust to place in the published findings.
• The problem lies in the fact that what we see in a research paper is the final, polished version of the work. It’s like buying a used car that has been meticulously cleaned and buffed for sale—you see the shiny exterior, but you don’t know the true history.
• The “history” of a scientific paper includes all the design choices, data decisions, and analyses that led to what was eventually published, and much of that is invisible to the reader.
• This is one reason why I strongly support pre-registration and registered reports. In these systems, researchers publicly document their study plan before data collection.
• This includes the research question, study design, sample size calculations, and planned statistical analyses. Once the study is complete, the published paper links back to the pre-registration, allowing readers to compare what was planned with what was actually done.
• In our car analogy, it’s like having a dashcam recording of the vehicle’s entire life—you can see how it was driven, maintained, and used, rather than just trusting the current presentation.
• Without pre-registration, there’s no single “magic” indicator that guarantees a paper’s reliability. However, one practical red flag to watch for is overconfidence in the conclusions.
• Sound scientific writing should present a balanced interpretation, acknowledging uncertainty and limitations. If a paper makes sweeping, definitive claims—especially from a single small study—that should raise suspicion.
• Conversely, when authors describe their results as tentative, call for further research, and avoid overstatement, it suggests a more responsible and credible approach to the evidence.

Meta analysis

42:18 -

• When it comes to meta-analyses, they are often considered the highest tier of evidence because they combine results from multiple studies, potentially overcoming some of the small sample size limitations in individual papers. In theory, this should give a clearer and more robust estimate of an effect.
• However, two major problems undermine their reliability in sports science. First, meta-analyses can only work with the studies that are available—and because of publication bias, those are often only the “successful” or statistically significant results.
• This means that any unpublished, non-significant studies remain invisible, skewing the combined effect size upward.
• Second, the quality of the meta-analysis itself matters greatly. Our recent review of meta-analyses in sports and exercise science—echoing earlier work by other colleagues—found that a large proportion contained statistical errors, sometimes as high as 80% in certain subfields like strength and conditioning.
• This points to a skill gap: many researchers in the field do not have the statistical expertise to conduct meta-analyses to the standard required for reliable conclusions.
• Because of these issues, I can’t simply recommend “go to the meta-analysis” as a safe shortcut to the truth. The better approach is to treat them as part of the picture but not the whole answer.
• Combine them with your own reading of key individual studies, and always consider the possibility of bias in both the underlying research and the synthesis.
• For coaches, athletes, and practitioners, this means accepting that sports science—especially in areas where effects are small—rarely gives neat, universally applicable answers.
• It’s not about discarding the science, but about using it alongside your own systematic self-experimentation. In practice, that means treating yourself or your athlete as a “sample size of one”: make small, careful changes; track the outcomes; and keep an open mind.
• When looking at new interventions or claims, a simple risk–benefit filter is useful. If the practice has no real risk of harm and is cheap or easy to try, it may be worth experimenting with, provided you evaluate it objectively.
• But if there is even a small risk of harm—and the evidence for benefit is inconsistent or weak—it’s usually best to avoid it. Ice baths are a good example: while they may offer a modest recovery benefit in certain contexts, there’s also evidence they can blunt training adaptations, so for most athletes the risk outweighs the potential gain.
• Ultimately, consume research with healthy skepticism, value the many scientists doing excellent and valuable work, and remember that in sport—as in much of science—what works best for you will often emerge from a blend of solid evidence, critical thinking, and thoughtful personal testing.

Issues in the research

49:55 -

• When two training interventions are fundamentally different—say, high-intensity 30/15 intervals versus steady-state moderate exercise—it makes intuitive and physiological sense that there could be meaningful differences in outcome.
• Even with a small sample size, you might reasonably expect to see a real effect, though the magnitude is likely exaggerated. In that case, the claim aligns with established training principles: higher-intensity work tends to produce certain adaptations more effectively than moderate continuous training.
• But when you’re comparing two nearly identical protocols—such as 30/15s versus 40/20s—the expected differences are minimal. A study showing a 5% performance advantage for one over the other, based on two groups of 10 athletes, should be treated with extreme caution.
• With such small sample sizes, random variation alone could easily explain the difference. Physiologically, the adaptations from these two interval formats would be very similar, so the “true” difference—if any—would likely be so small that detecting it reliably would require a far larger study.
• This is where it’s important not to get lost in the fine details and lose sight of the bigger picture. The body adapts primarily to the overall stress and load placed upon it, not to tiny tweaks in interval structure.
• Strength and conditioning literature offers a parallel example: debates over whether 60 seconds or 75 seconds of rest between sets is “optimal” often ignore the fact that both rest periods deliver a similar training stimulus in the context of a well-structured program.
• In reality, there may be no single “best” option. Slightly different methods can produce similar results, and the best approach might involve incorporating both for variety, maintaining adherence, and ensuring a broad range of stimuli.
• In many cases, obsessing over marginal differences risks missing the much larger performance gains that come from consistently applying the basic principles of progressive overload, recovery, and specificity.

Practical tips for the average athlete

51:42 -

• First, understand that p-values are not a measure of effect size or strength of evidence. A p-value simply tells you the probability that you would see an effect at least as large as the one measured if there were actually no true effect, given the data and statistical model. It says nothing about how big or practically meaningful the effect is.
• Instead, focus on effect sizes. These come in two forms:
  • Unstandardised effect sizes, which are in the original measurement units (e.g., “the shaved-leg group finished the time trial 7 seconds faster on average”). These are very easy to interpret practically.
  • Standardised effect sizes, such as Cohen’s d, which adjust the raw difference by the variability in the sample. This makes them comparable across studies using different measurement units. A Cohen’s d of 1.0 means the difference between groups is equal to one standard deviation of the data; 0.5 is about half a standard deviation; 0.2 is about one-fifth.
• The nuance is that the practical meaning of a given effect size depends on the context.
• For something that is easy, cheap, and risk-free—like drinking a coffee before a race—even a small effect size (e.g., 0.15) can be valuable.
• For something that requires large time, cost, or risk investments—like an 8-week strength-training block—an effect size that small might not justify the effort if your only goal is a marginal gain in a specific outcome.
• Sample size also interacts with effect size.
• If the expected effect is small (as with caffeine), you need a larger sample to be confident it exists at all. A small sample producing a small effect size should raise red flags.
• If the expected effect is large (as in comparing trained vs. untrained groups on a big physiological change), you can be reasonably confident with a smaller sample—though more participants always improve precision.
• This means the right way to interpret a study is to ask: 
  • What is being measured?
  • How big is the reported effect, and in what units?
  • How variable are the measurements (standard deviation)?
  • Is the sample size appropriate for detecting an effect of this magnitude?
• It’s not as simple as “big sample = good” or “p < 0.05 = true.” You have to weigh sample size, expected effect magnitude, and variability together—and then ask if the result is practically meaningful in the real-world context of the sport or training decision you’re making.

Joe’s research

59:48 -

• There’s no doubt that research practices in sports science are improving, and a few key initiatives are leading the way.
• One major step forward has been mandatory sample size justification in certain journals. This forces researchers to show they have planned their study with enough statistical power to meaningfully detect the effect they’re investigating. It’s a safeguard against the small, underpowered studies that have historically contributed to inflated claims and poor reproducibility.
• Another important change is mandatory pre-registration. Here, researchers publicly register their study plan—including hypotheses, methods, and statistical analysis—before collecting any data. This makes the process transparent and reduces the risk of “data dredging” or selectively reporting results that look more interesting.
• We’re also seeing growing adoption of open science practices—making data, analysis code, and sometimes even study materials freely available.
• This allows other scientists to verify findings, run alternative analyses, or build on the work without starting from scratch. Open data fosters cumulative science rather than siloed projects.
• On the technological side, there’s promising work from people like Daniel Lakens—a leading figure in meta-science—who is developing Paper Check, an AI-driven tool that can scan research papers for methodological and reporting issues.
• Initially aimed at helping academics check their work before submission, tools like this could eventually help practitioners, journalists, and non-academics get an objective quality snapshot of a paper before drawing conclusions from it.

Takehome messages

1:02:08 -

• For listeners who approach research from the consumer side rather than the academic side, the key is not to be discouraged by the recent criticisms you might hear—whether in sports science or science more broadly. The vast majority of scientists are well-intentioned, working hard to produce quality research, and many are making important contributions that genuinely move the field forward.
• The most important point is that no single study should ever be taken as definitive proof of anything. One paper is a single data point in a much bigger picture. Whenever possible, look at the broader body of work before making decisions or changing your practice. Patterns across multiple studies are far more reliable than isolated findings.
• This is not an anti-science message—far from it. I am a scientist myself, and I work with colleagues who are producing rigorous, valuable work every day. The challenges we’ve discussed stem from systemic issues in the research ecosystem, not from widespread deliberate misconduct.
• The encouraging news is that things are changing. Open science practices, better methodological standards, and emerging tools such as AI-assisted paper reviews are already making an impact. If these trends continue, we can expect the credibility and transparency of sports science to improve rapidly in the coming years.

Sports and non-sporting recommendations in Ireland

1:04:04 -

• For a sporting recommendation in Ireland, I’d suggest the Art O’Neill Challenge—a unique and historic ultra event in Dublin. It commemorates a real escape in 1592, when three inmates fled Dublin Castle in the middle of a winter night, covering roughly 60 kilometres to reach the remote Glenmalure valley in County Wicklow.
• The modern version starts at night in the heart of Dublin and is largely self-navigated, taking you from the urban streets into the rugged Wicklow Mountains. It’s notoriously hard to secure a place, but it’s one of the most distinctive endurance events in the country.
• For a non-sporting recommendation, explore Donegal in Ireland’s northwest. The region offers some of the most spectacular natural scenery in the country—wild coastlines, dramatic cliffs, and sweeping mountain landscapes. Hiking here is unforgettable, and the remote beauty of the area makes it feel far removed from the busier tourist spots.

LINKS AND RESOURCES

• Joe's ResearchGate profile

• Sports Science Replication Centre website 

• Replication concerns in sports and exercise science: a narrative review of selected methodological issues in the field - Mesquida et al. 2022

• Estimating the Replicability of Sports and Exercise Science Research - Murphy et al. 2025

• Publication bias, statistical power and reporting practices in the Journal of Sports Sciences: potential barriers to replicability - Mesquida et al. 2023

• On the replicability of sports and exercise science research: assessing the prevalence of publication bias and studies with underpowered designs by a z-curve analysis - Mesquida et al. 2025

• A scoping review of the transparency, reporting practices and methodological rigor of meta-analyses published in Sports Medicine - Mesquida et al. 2025

• prevalence, reporting practices, and methodological quality of a priori power analyses in sports and exercise science research - Mesquida et al. 2025