Replication Studies in Ecology and Evolution: A Guide to Design and Interpretation

A replication study in ecology and evolution repeats or extends prior research to test whether findings hold under new conditions, with new samples, or with improved methods. For students, researchers, and life-science professionals, a replication study is a deliberate research design that asks whether an observed effect is robust, generalizable, or conditional on specific biological or environmental contexts. This guide explains what replication means in ecological and evolutionary research, how to design a replication study that matches the scale of your inference, how to interpret results that confirm or contradict earlier work, and how to document methods so others can repeat your efforts.

What Counts as a Replication Study in Ecology and Evolution

A replication study in ecology and evolution is an empirical investigation that repeats a previously reported experiment, observation, or analysis to determine whether the original result can be reproduced. The term covers several distinct approaches. A direct replication attempts to repeat the original study as closely as possible, using the same species, methods, and conditions. A conceptual replication tests the same hypothesis or question but uses different species, systems, or methods to see whether the underlying idea holds more broadly. A quasi-replication, sometimes called a partial replication, repeats part of the original design or applies it to a new context without matching every detail.

The distinction matters because ecological and evolutionary systems are inherently variable across space and time. A finding that holds for one population of oysters in one estuary may not hold for another population in a different region. A result that appears in one breeding season may disappear the next. Replication studies are the primary tool for determining whether a finding is a general biological principle or a local artifact of particular conditions.

The Principles of experimental design for ecology and evolution paper emphasizes that many contemporary studies fail to replicate at the appropriate biological or organizational level, so causal inference may have less support than often assumed. The author argues that experimental design should be discussed in terms of the scale of replication relative to the scale at which inferences are sought. In practical terms, if you want to make a claim about all populations of a species, your replication must include multiple populations. If you want to make a claim about a single population, replication within that population may be sufficient.

Replication is also a matter of scientific culture. A survey of ecologists found that 97% of respondents considered replication studies important, 91% thought they were not prevalent enough, and 62% believed they were suitable for publication in all journals. Yet the actual prevalence of direct replication studies in the ecology literature is far lower than researchers estimate. The same survey identified obstacles including the difficulty of conducting replication studies and the challenges of funding and publishing them. The cost-benefit analysis of replication similarly notes that replication within species and systems is troublingly rare and that the current incentive structure needs to change if ecologists and evolutionary biologists are to value replication sufficiently.

Why Replication Is Difficult in Ecological and Evolutionary Systems

Ecological and evolutionary systems present challenges that are less severe in other scientific fields. Laboratory-based disciplines can often control environmental conditions precisely and use genetically identical organisms. Ecology and evolution frequently study organisms in the field, where temperature, rainfall, predation, competition, and countless other factors vary beyond investigator control.

The discussion of naturally clonal vertebrates highlights one solution to this problem. Researchers have developed cloned or isogenic model organisms because replication requires genetic uniformity. However, many researchers are reluctant to use traditional animal model systems for ecology and evolution questions because of concerns about relevance or inbreeding. The authors point out that a substantial number of vertebrate species reproduce clonally in nature, and these naturally evolved, phenotypically complex animals can push the boundaries of traditional experimental design.

Even when genetic variation is controlled, environmental variation remains. A study of green algae-induced oyster microbiota dysbiosis illustrates the problem. Previous research showed that Pacific oysters exposed to green algae from the genus Ulva were more susceptible to infection by Ostreid herpesvirus type 1, coinciding with dysbiosis of the bacterial microbiota. Because the microbiota of macroalgae and oysters vary with time and geographical location, the researchers conducted an experiment with macroalgae and oysters from several origins to test whether the previously observed interaction could be extended more generally. They found that green algae increased mortality risk regardless of geographical origin or oyster line, with one exception involving algae from the English Channel. This study demonstrates both the value and the complexity of replication across spatial scales.

Temporal variation is equally important. Ecological systems change across seasons, years, and longer cycles. A replication study conducted in a different year may produce different results because of weather patterns, population cycles, or community composition changes that have nothing to do with the validity of the original finding. Researchers must decide whether these differences represent a failure to replicate or an important biological insight about context dependence.

The Scale of Replication Must Match the Scale of Inference

The most common design error in ecological and evolutionary replication is a mismatch between the scale of replication and the scale of inference. The Principles of experimental design for ecology and evolution paper provides a guide for identifying the appropriate scale of replication for common experimental designs. The core principle is that you cannot make inferences about a level of biological organization that you did not replicate.

Consider a study that measures the effect of a pollutant on growth rate in a single population of fish. If the researcher wants to conclude that the pollutant harms fish of that species generally, the study must include multiple populations. If the researcher only wants to conclude that the pollutant harms fish in that particular lake, replication within the lake may be adequate. The same logic applies to temporal scales. A study conducted in one year supports inferences about that year. Inferences about long-term patterns require replication across years.

The paper also discusses the merits of replicating multiple scales of biological organization simultaneously. A well-designed study might include multiple populations, multiple sites within each population, and multiple individuals within each site. This hierarchical approach allows the researcher to partition variation among scales and determine where the treatment effect operates.

Practical guidance for matching scale to inference includes the following steps. First, state the target population for your inference clearly before designing the study. Second, identify the sources of biological and environmental variation that could affect the outcome. Third, choose replication units that capture that variation. Fourth, ensure that the number of replicates at each level is sufficient to detect the effect size you consider biologically meaningful. Fifth, document the rationale for your scale choices so reviewers and readers can evaluate whether your design supports your conclusions.

Direct Replication Versus Conceptual Replication

Direct replication attempts to reproduce the original study as closely as possible. The goal is to determine whether the original result is reproducible under the same or very similar conditions. Direct replication is common in fields like psychology, where the replication study of mental imagery repeated a highly cited but underpowered 2007 study with much larger samples. The original study had only eight participants and found that more vivid imagery increased interference between imagined and perceptual content. The replication recruited 185 online participants in one experiment and 56 participants in another, including individuals with aphantasia and hyperphantasia. The researchers were unable to replicate the original effect and instead observed performance benefits for color-word congruency across the mental imagery spectrum.

Direct replication in ecology and evolution faces practical obstacles. Field conditions cannot be exactly reproduced. Populations change over time. Even in controlled environments, subtle differences in husbandry, diet, or microbial communities can affect outcomes. The oyster microbiota study is a useful example because it attempted to extend previous findings across geographical origins and oyster lines, which is a form of conceptual replication instead of direct replication.

Conceptual replication tests whether the underlying hypothesis holds when the specific system or methods change. The Spanish verbal mood replication is an example from linguistics. The original study examined how English-speaking learners of Spanish interpreted verbal moods in adverbial clauses. The replication tested whether the results extended to Swedish and French learners of Spanish. The researchers found that multiple factors influenced variable interpretation of verbal moods and that there were differences between the French and Swedish groups. This study enhanced the confirmatory power of some of the original findings while suggesting that the learners' first language leads to diverging findings.

For ecology and evolution, conceptual replication often means testing a hypothesis in a different species, ecosystem, or geographical region. A finding about competition in temperate grasslands might be tested in tropical savannas. A finding about predator-prey dynamics in lakes might be tested in marine systems. Conceptual replication is valuable because it tests the generality of ecological and evolutionary principles, but it cannot determine whether the original study was methodologically sound. Only direct replication can do that.

Designing a Replication Study

Designing a replication study requires the same rigor as designing any other empirical study, with additional attention to the relationship between the original study and the replication. The Experimental Design Assistant from the NC3Rs provides a tool for planning and visualizing experimental designs, including randomization, blinding, and replication considerations. While developed primarily for animal research, its principles apply broadly to ecological and evolutionary studies.

The first step is to define the research question precisely. What specific claim from the original study are you testing? What outcome measure will you use? What effect size would you consider meaningful? The coral reef restoration framework emphasizes that whether interventions are considered successful depends on whether goals and outcomes were clearly defined before implementing the intervention. The same principle applies to replication studies. You must specify your success criteria before you begin.

The second step is to determine the appropriate type of replication. If you are testing whether the original finding is reproducible under the same conditions, use a direct replication. If you are testing whether the finding generalizes to other systems, use a conceptual replication. If you are testing whether the finding holds in a related but different context, use a quasi-replication.

The third step is to determine the scale of replication. Following the Principles of experimental design for ecology and evolution, you must replicate at the level at which you seek inference. If you want to make claims about multiple populations, include multiple populations. If you want to make claims about multiple years, include multiple years. If you want to make claims about a single population in a single year, replicate within that population and year.

The fourth step is to document your methods thoroughly. The Research Data Framework from the National Institute of Standards and Technology provides guidance on data management practices that support reproducibility. The EQUATOR Network provides reporting guidelines for health research that emphasize transparent and complete reporting of methods. While these resources are not specific to ecology and evolution, their principles apply. Your methods section should include enough detail that another researcher could repeat your study without contacting you for clarification.

The fifth step is to plan your analysis before you collect data. How will you determine whether your results replicate the original finding? Will you use statistical significance, effect size comparison, or confidence interval overlap? The discussion of replication methods notes that there are multiple ways to conduct a replication study beyond statistical significance. You should specify your criteria in advance to avoid the temptation to interpret ambiguous results in a way that supports your expectations.

At a Glance: Replication Study Types and Design Choices

The table below summarizes the main replication approaches, their purposes, and the design considerations that apply to each.

A second useful table addresses the documentation categories that support replication.

Documenting Environmental Conditions and Methods

Documentation is the foundation of replication. Without detailed records of environmental conditions and methods, other researchers cannot repeat your study, and you cannot interpret differences between your results and the original findings. The Research Data Framework provides a structure for managing research data throughout its lifecycle, from planning through collection, analysis, and sharing.

For ecological and evolutionary replication studies, documentation should include the following categories. First, the study site or system, including geographical coordinates, habitat type, and any relevant historical information. Second, the study organism, including species, population, genetic background if known, and source. Third, environmental conditions during the study, including temperature, precipitation, photoperiod, and any unusual events such as storms, droughts, or disease outbreaks. Fourth, the experimental protocol, including treatments, controls, randomization procedures, and blinding. Fifth, the measurement methods, including instruments, calibration, and quality control procedures. Sixth, the data analysis methods, including statistical models, software, and code.

The oyster microbiota study provides a model for documenting environmental conditions. The researchers specified the sources of macroalgae and oysters, the acclimation procedures, the challenge protocol, and the methods for following mortality and virus replication dynamics. This level of detail allowed them to identify the one exception to their general finding and to discuss possible explanations.

A practical checklist for documenting environmental conditions includes the following items. Record the exact dates of the study and any seasonal context. Record weather conditions, including temperature ranges, rainfall, and any extreme events. Record the condition of the study organisms, including health status, size, age, and reproductive state. Record the physical and chemical properties of the environment, such as water quality parameters for aquatic studies or soil properties for terrestrial studies. Record any disturbances or management activities that occurred during the study period. Record the names and versions of all software used for data collection and analysis. Record any deviations from the original protocol and the reasons for those deviations.

Interpreting Replication Results

Interpreting replication results requires more than comparing p-values. A replication that fails to reach statistical significance does not necessarily mean the original finding was wrong. The replication may have been underpowered, the conditions may have differed in important ways, or the original finding may have been a false positive. Conversely, a replication that reaches statistical significance does not necessarily confirm the original finding. The replication may have shared the same methodological flaws as the original study.

The methotrexate response replication study provides an example of rigorous interpretation. The researchers replicated a selection of 25 single nucleotide polymorphisms associated with response to methotrexate in patients with rheumatoid arthritis. They considered consistency between outcomes, p-values accounting for the number of SNPs, and independence from potential confounders. Only one SNP fulfilled the high association standards. This approach demonstrates the importance of multiple criteria for interpreting replication results.

The Jadad algorithm replication study illustrates another challenge in interpretation. The researchers attempted to replicate author assessments using the Jadad algorithm to choose the best systematic review. They found that in 62% of cases, they were unable to replicate the Jadad assessment and ultimately chose a different systematic review than the authors. However, 86% of independent Jadad assessments agreed in direction of the findings. This study shows that even when the final choice differs, the underlying evidence may point in the same direction.

For ecological and evolutionary replication studies, interpretation should consider the following factors. First, the effect size and its confidence interval. A replication that produces a similar effect size with a wider confidence interval provides more support than one that produces a smaller effect size with a narrow confidence interval. Second, the direction of the effect. A replication that produces an effect in the same direction but with a larger p-value may be consistent with the original finding. Third, the biological context. Differences in environmental conditions, population history, or community composition may explain differences in results. Fourth, the methodological quality of both the original study and the replication. A replication that improves on the original design may produce different results because it corrects flaws in the original.

The spatio-temporal model of influenza A virus infection demonstrates the value of modeling for interpretation. The researchers developed a spatially explicit, stochastic model that integrates virus and defective interfering particle replication, interferon signaling, and alternative dispersal modes. The model captured experimentally observed plaque morphologies and showed that interferon production peaks at an intermediate defective interfering particle ratio. Models like this can help researchers understand why replication studies produce different results by exploring the conditions under which the original finding holds.

Common Failure Patterns in Replication Studies

Replication studies in ecology and evolution fail for predictable reasons. Understanding these failure patterns can help researchers design better studies and interpret results more accurately.

The first failure pattern is underpowered replication. A replication study that uses too few replicates cannot detect the original effect even if it is real. The mental imagery replication explicitly addressed this problem by recruiting much larger samples than the original study. In ecology and evolution, power analysis should be conducted before data collection to determine the number of replicates needed to detect the effect size of interest.

The second failure pattern is inadequate documentation. If the original study did not document environmental conditions, methods, and analysis procedures in sufficient detail, the replication cannot match the original conditions. The Research Data Framework and the EQUATOR Network provide guidance for improving documentation practices.

The third failure pattern is scale mismatch. A replication that uses a different scale of replication than the original study may produce different results for reasons unrelated to the validity of the original finding. The Principles of experimental design for ecology and evolution emphasizes that the scale of replication must match the scale of inference.

The fourth failure pattern is publication bias. Replication studies that fail to confirm the original finding may be difficult to publish, while replication studies that confirm the original finding may be more likely to be published. This bias distorts the scientific record and makes it difficult to assess the true reproducibility of ecological and evolutionary research.

The fifth failure pattern is confirmation bias. Researchers who expect to replicate the original finding may unconsciously design their study, analyze their data, or interpret their results in ways that favor confirmation. Blinding, pre-registration, and pre-specified analysis plans can reduce this bias.

The sixth failure pattern is ignoring context dependence. Ecological and evolutionary findings are often conditional on specific environmental or biological contexts. A replication that fails because the conditions differ from the original study may be misinterpreted as a failure of the original finding when it actually demonstrates important context dependence. The biotic interactions biogeography framework discusses how species interactions shape biodiversity patterns across scales and emphasizes the importance of considering environmental gradients and spatial scales.

Records and Measurements for Replication Studies

Maintaining detailed records is essential for replication studies. The records serve two purposes. First, they allow other researchers to repeat your study. Second, they allow you to interpret differences between your results and the original findings.

The Research Data Framework provides a structure for managing research data throughout its lifecycle. Key elements include data planning, data collection, data documentation, data quality assurance, data analysis, and data sharing. For replication studies, data documentation is particularly important because the value of the replication depends on the ability of others to understand what was done and why.

A practical records system for a replication study includes the following components. First, a study protocol that describes the research question, the replication type, the scale of replication, the methods, and the analysis plan. Second, a field or laboratory notebook that records daily activities, observations, and any deviations from the protocol. Third, raw data files that are stored in a stable format and backed up regularly. Fourth, metadata that describes the data files, including variable names, units, and coding schemes. Fifth, analysis scripts that document the statistical methods and can be rerun by others. Sixth, a results log that records the outcomes of analyses and any decisions made during the analysis process.

The NCBI Literature Resources and PubMed provide access to the scientific literature that replication studies build upon. Researchers should search these databases thoroughly to identify all relevant prior studies, including those that may have attempted replication before. The NCBI Literature Resources also provides tools for finding related articles and tracking citations, which can help researchers identify studies that have replicated or failed to replicate a particular finding.

Quality Controls and Welfare Considerations

Quality controls in replication studies serve two purposes. They ensure that the data are reliable, and they ensure that the study organisms are treated ethically. The Experimental Design Assistant from the NC3Rs provides guidance on experimental design that incorporates both scientific and welfare considerations.

For studies involving animals, welfare considerations are paramount. The Experimental Design Assistant helps researchers plan experiments that minimize animal suffering while maximizing scientific value. Key principles include using the minimum number of animals needed to achieve the scientific objective, avoiding unnecessary replication, and monitoring animal welfare throughout the study.

For field studies, welfare considerations extend to the ecosystem. Researchers should minimize disturbance to study sites and populations. They should also consider the potential impacts of their study on other species and on the broader ecosystem.

Quality controls for data collection include calibration of instruments, standardization of measurement procedures, and regular checks for data errors. The coral reef restoration framework emphasizes the importance of distinguishing between metrics that quantify production and outplanting efforts and metrics that measure recovery of community structure and ecosystem functioning. This distinction is relevant to replication studies, which must measure the same outcomes as the original study to be comparable.

The spatio-temporal model of influenza A virus infection demonstrates the value of quality controls in modeling studies. The model is available as an interactive platform, allowing other researchers to explore the model behavior and test alternative assumptions. This transparency is a form of quality control that supports replication.

Limitations of Replication Studies

Replication studies have inherent limitations that researchers must acknowledge. The most fundamental limitation is that replication cannot prove a finding is true. A successful replication increases confidence in the original finding, but it cannot rule out the possibility that both the original study and the replication share a common flaw. Conversely, a failed replication does not prove the original finding is false. The replication may have differed from the original in important ways, or the original finding may be context-dependent.

The discussion of replication in ecology and evolution notes that replication within species and systems is troublingly rare, and even quasi-replications in different systems are often insufficient. This scarcity means that the evidence base for many ecological and evolutionary findings is weaker than it appears. The authors argue that the current incentive structure needs to change if ecologists and evolutionary biologists are to value scientific replication sufficiently.

Another limitation is that replication studies are often more difficult to fund and publish than novel studies. The survey of ecologists found that while most ecologists considered replication studies important and worth funding, the actual prevalence of direct replication studies in the literature is much lower than researchers estimate. This disconnect between attitudes and practice reflects the structural obstacles to replication.

The Baltimore Classification of Viruses provides an example of a conceptual framework that has been tested and refined over decades. The classification of viruses by routes of genome expression has remained an integral part of the conceptual foundation of biology, but it has been extended and modified as new evidence emerged. This history illustrates the value of replication and extension in building robust scientific knowledge.

The Bacteriophage T4 genome similarly demonstrates the importance of replication in molecular biology. The complete genome sequence of phage T4 has provided countless contributions to the paradigms of genetics and biochemistry, and the redundancy of DNA replication and recombination systems reveals how phage and other genomes are stably replicated and repaired in different environments. This knowledge provides insight into genome evolution and adaptations to new hosts and growth environments.

Professional Escalation Criteria

Researchers conducting replication studies should know when to escalate concerns to supervisors, collaborators, or institutional authorities. The following situations warrant escalation.

First, if the replication study produces results that contradict the original finding in a way that could have significant scientific or practical consequences, escalate to the research team and consider whether the discrepancy should be reported. Second, if the replication study reveals evidence of fraud, data fabrication, or other research misconduct in the original study, escalate to the institutional research integrity office. Third, if the replication study identifies a safety concern, such as an unexpected hazard to researchers or study organisms, escalate to the institutional safety office. Fourth, if the replication study reveals a welfare concern, such as unexpected suffering of study animals, escalate to the institutional animal care and use committee. Fifth, if the replication study produces results that could affect public health, environmental policy, or other decisions with broad consequences, escalate to the appropriate authorities.

The EQUATOR Network provides reporting guidelines that can help researchers decide what information to include in their reports and how to present it transparently. The NCBI Literature Resources and PubMed provide access to the literature that can help researchers interpret their results in context.

Frequently Asked Questions

What is the difference between a direct replication and a conceptual replication?

A direct replication repeats the original study as closely as possible, using the same species, methods, and conditions. A conceptual replication tests the same hypothesis or question but uses different species, systems, or methods. Direct replication tests whether the original result is reproducible. Conceptual replication tests whether the underlying idea holds more broadly. Both are valuable, but they answer different questions.

How many replicates do I need for a replication study?

The number of replicates depends on the effect size you want to detect, the variability in your system, and your chosen significance level and power. Conduct a power analysis before collecting data. The Principles of experimental design for ecology and evolution emphasizes that the scale of replication must match the scale of inference. If you want to make claims about multiple populations, you need multiple populations, beyond multiple individuals within one population.

What should I do if my replication study fails to confirm the original finding?

First, check whether your study was adequately powered to detect the original effect. Second, compare your methods and conditions to the original study to identify any differences. Third, consider whether the original finding may be context-dependent. Fourth, report your results honestly, including the details of your methods and conditions. A failed replication is a valuable scientific contribution, especially if you can identify why the result differed.

How do I document environmental conditions for a replication study?

Record the exact dates of the study, weather conditions, the condition of study organisms, the physical and chemical properties of the environment, any disturbances or management activities, and any deviations from the original protocol. The Research Data Framework provides guidance on data management practices that support reproducibility.

Can I publish a replication study in an ecology or evolution journal?

Most ecologists consider replication studies suitable for publication in all journals, according to the survey of ecologists. However, the actual prevalence of direct replication studies in the literature is much lower than researchers estimate. You may need to explain the value of your replication study clearly in your cover letter and emphasize what it adds beyond the original study.

What is the difference between replication and pseudo-replication?

Replication involves independent units that capture the variation at the scale of inference. Pseudo-replication occurs when the analysis treats non-independent units as independent, leading to inflated confidence in the results. The Principles of experimental design for ecology and evolution discusses the importance of identifying the appropriate scale of replication and avoiding pseudo-replication.

How do I interpret a replication study that produces a different effect size but the same direction of effect?

A replication that produces an effect in the same direction but with a different magnitude may be consistent with the original finding. Consider the confidence intervals around both effect sizes. If they overlap substantially, the difference may be due to sampling variation. If they do not overlap, the difference may reflect context dependence or methodological differences.

What resources are available to help me design a replication study?

The Experimental Design Assistant from the NC3Rs provides a tool for planning and visualizing experimental designs. The Research Data Framework provides guidance on data management. The EQUATOR Network provides reporting guidelines. The NCBI Literature Resources and PubMed provide access to the scientific literature.

Related Articles

• RT-qPCR Experiments: Design Choices That Affect Interpretation
• Single-Nucleus RNA Sequencing: Study Design, Workflow, and Interpretation
• Single-Nucleus RNA Sequencing: Study Design, Workflow, and Interpretation
• Single-Nucleus RNA Sequencing: Study Design, Workflow, and Interpretation
• ELISA Workflow: Designing Controls and Interpreting Plate Results

References and Further Reading

• Research Data Framework. National Institute of Standards and Technology.
• EQUATOR Network. EQUATOR Network.
• Experimental Design Assistant. NC3Rs.
• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.
• Principles of experimental design for ecology and evolution.. Ecology letters, 2024.
• The Baltimore Classification of Viruses 50 Years Later: How Does It Stand in the Light of Virus Evolution?. Microbiology and molecular biology reviews : MMBR, 2021.
• Bacteriophage T4 genome.. Microbiology and molecular biology reviews : MMBR, 2003.
• Replicating research in ecology and evolution: feasibility, incentives, and the cost-benefit conundrum.. BMC biology, 2015.
• The role of replication studies in ecology.. Ecology and evolution, 2020.
• Sexual conflict.. Current biology : CB, 2019.
• Naturally clonal vertebrates are an untapped resource in ecology and evolution research.. Nature ecology & evolution, 2019.
• Ecology and Evolution in the RNA World Dynamics and Stability of Prebiotic Replicator Systems.. Life (Basel, Switzerland), 2017.
• Biotic interactions biogeography: A framework for understanding how species interactions shape biodiversity patterns across scales.. 2026.
• Towards a systematic framework to assess restoration success of interventions in coral reef ecosystems.. 2026.
• Spatial and temporal consistency in green algae-induced oyster microbiota dysbiosis and associated increased disease risk.. 2026.
• Spatio-temporal modelling of in vitro influenza A virus infection: The impact of defective interfering particles on the type I interferon response.. 2026.
• Individual variability in mental imagery vividness does not predict perceptual interference with imagery: A replication study of Cui et al. (2007).. Journal of experimental psychology. General, 2025.
• The interpretation of verbal moods in Spanish: A close replication of Kanwit and Geeslin (2014). Studies in Second Language Acquisition, 2024.
• How can clinicians choose between conflicting and discordant systematic reviews? A replication study of the Jadad algorithm. BMC Medical Research Methodology, 2021.
• A Survey of Using YouTube as Supplementary Material with University English Language Learners in Vietnam: A Replication Study. Rupkatha Journal on Interdisciplinary Studies in Humanities, 2021.
• There's more than one way to conduct a replication study: Beyond statistical significance.. Psychological methods, 2016.
• Replication study of polymorphisms associated with response to methotrexate in patients with rheumatoid arthritis. Scientific Reports, 2018.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.