# Cohen's d vs. eta-squared vs. odds ratio ## Key Takeaways - **Effect size metrics must align with study design and outcome type:** Cohen's d is appropriate for continuous outcomes in two-group comparisons (e.g., treated vs. control cells), eta-squared for continuous outcomes in ANOVA designs with three or more groups (e.g., multiple treatment doses), and odds ratios for categorical outcomes in contingency tables (e.g., infection status in exposed vs. unexposed populations). - **Standardized mean differences (Cohen's d) quantify effect magnitude in units of pooled standard deviation**, allowing comparison across studies with different measurement scales; interpretation benchmarks (0.2 small, 0.5 medium, 0.8 large) are context-dependent in biological research. - **Eta-squared represents the proportion of total variance in a continuous outcome explained by a categorical predictor**, making it suitable for ANOVA designs comparing multiple groups (e.g., genotypes); partial eta-squared is preferred in factorial designs to isolate the effect of a specific factor. - **Odds ratios compare the odds of an event occurring between two groups for categorical outcomes**, commonly used in case-control studies or logistic regression to assess risk (e.g., mutation presence in cases vs. controls); interpretation requires careful consideration of the event's prevalence to avoid overestimating relative risk. - **Reporting effect sizes alongside confidence intervals is critical for assessing the precision and biological significance of findings**, independent of sample size, and is increasingly mandated by reporting guidelines like those from the EQUATOR Network. - **A structured, study-level effect size reporting log is essential for documenting the rationale behind metric selection, calculation details, and interpretation**, ensuring reproducibility and facilitating meta-analysis by providing an audit trail connecting design, test, metric, value, and rationale. --- ## Quick Answer - Choose Cohen's d for two-group mean comparisons from t-tests, eta-squared for ANOVA designs with three or more groups, and odds ratios for categorical outcomes in contingency tables. - Match the effect size metric to your study design and statistical test, then report it alongside confidence intervals to communicate magnitude and precision. - No single effect size works for every design, and converting between metrics requires assumptions that may not hold in biological data. ## Understanding Effect Size in Biological Research Statistical significance testing answers whether an observed difference or association could plausibly arise by chance. Effect size answers a different question: how large is the difference or association in meaningful terms? In biological research, a statistically significant result with a trivial effect size can mislead interpretation, while a biologically important effect may fail to reach significance in an underpowered study. Reporting effect sizes allows readers to judge practical importance independent of sample size. The choice among Cohen's d, eta-squared, and odds ratio depends on the structure of the data and the statistical test applied. Cohen's d suits continuous outcomes compared between two groups. Eta-squared suits analysis of variance designs where the outcome is continuous and predictors are categorical. Odds ratio suits categorical outcomes arranged in contingency tables. Each metric answers a distinct question about the data, and each has its own interpretation scale and reporting conventions. Biological data present particular challenges. Measurements often violate normality assumptions, variance differs between groups, and sample sizes are constrained by cost and ethics. These conditions affect which effect size is appropriate and how it should be interpreted. Researchers who understand the assumptions behind each metric can select the one that matches their design and communicate results accurately to readers. ## Core Principles of Effect Size Metrics ### What Effect Size Measures Effect size quantifies the magnitude of an experimental effect independent of sample size. A p-value combines effect size with sample size and variability, so a large sample can produce a small p-value for a negligible effect. Effect size separates the magnitude from the sample size, allowing comparison across studies with different numbers of observations. Three families of effect sizes exist. The d family describes standardized mean differences between groups. The r family describes strength of association between variables. The odds ratio family describes the ratio of odds in categorical data. Each family connects to specific statistical tests and study designs. ### Why Reporting Effect Size Matters Reporting effect sizes supports meta-analysis, power analysis, and interpretation of biological importance. Without effect sizes, readers cannot judge whether a statistically significant result has practical relevance. Journals increasingly require effect sizes in addition to p-values, and reporting guidelines emphasize their inclusion. The [EQUATOR Network](https://www.equator-network.org/) provides reporting guidelines that help researchers select appropriate statistical reporting practices. These guidelines cover the information that should appear in a research report, including effect sizes and their precision. Following such guidelines improves transparency and allows readers to assess the strength of evidence. ### The Relationship Between Effect Size and Study Design Study design determines which effect size is appropriate. A two-group comparison with a continuous outcome calls for Cohen's d. A multi-group comparison with a continuous outcome calls for eta-squared. A categorical outcome with two or more groups calls for odds ratio. Using the wrong metric for the design produces numbers that are difficult to interpret or that answer a different question than the one posed. The decision framework begins with identifying the outcome variable type and the number of groups being compared. Continuous outcomes with two groups lead to Cohen's d. Continuous outcomes with three or more groups lead to eta-squared. Categorical outcomes lead to odds ratio. This framework covers the most common designs in biological research. ## Cohen's d for Two-Group Comparisons ### Definition and Calculation Cohen's d is a standardized mean difference. It expresses the difference between two group means in units of the pooled standard deviation. A d of 0.5 means the two group means differ by half a standard deviation. This standardization allows comparison across studies that use different measurement scales. The calculation uses the means of the two groups, the sample sizes, and the standard deviations. The pooled standard deviation weights each group's variance by its sample size. The resulting value is unitless, which permits comparison across different outcome measures. ### Interpretation Benchmarks Cohen proposed conventional benchmarks for interpreting d values. A d of 0.2 is considered small, 0.5 medium, and 0.8 large. These benchmarks are arbitrary and should be interpreted in the context of the biological question. A small effect on a clinically important outcome may be more meaningful than a large effect on a trivial outcome. In biological research, the context determines whether a given d is meaningful. A d of 0.3 in a study of enzyme activity may be biologically important if the enzyme is a drug target. The same d in a study of a behavioral outcome may be trivial. Researchers should interpret d in the context of the specific system and the potential consequences of the effect. ### When to Use Cohen's d Cohen's d is appropriate when the study compares two groups on a continuous outcome. Common examples include comparing treated versus untreated cells, wild-type versus knockout animals, or two experimental conditions. The t-test is the corresponding statistical test, and d is the effect size that accompanies it. Cohen's d is also useful for meta-analysis because it standardizes results across studies that use different measurement scales. A meta-analysis can combine d values from multiple studies to estimate the overall effect. This requires that the studies use similar designs and that the effect sizes are comparable. ### Limitations of Cohen's d Cohen's d assumes the two groups have similar variances. When variances differ substantially, the pooled standard deviation may not represent either group well. In such cases, a version of d that uses the control group's standard deviation may be more appropriate. Cohen's d does not convey the direction of the effect unless the sign is reported. A negative d indicates the second group has a higher mean than the first. Researchers should report the sign and the direction of the effect to avoid ambiguity. ## Eta-Squared for ANOVA Designs ### Definition and Calculation Eta-squared is a measure of association for analysis of variance designs. It represents the proportion of total variance in the outcome that is attributable to the grouping variable. An eta-squared of 0.3 means 30 percent of the variance in the outcome is explained by group membership. The calculation divides the sum of squares between groups by the total sum of squares. The resulting value ranges from 0 to 1, with higher values indicating stronger association. Eta-squared is a descriptive measure that does not adjust for the number of predictors or the sample size. ### Partial Eta-Squared Partial eta-squared is a variant that measures the proportion of variance explained by a factor after removing the variance explained by other factors. It is commonly reported in factorial ANOVA designs with multiple independent variables. Partial eta-squared can be larger than eta-squared because it uses a different denominator. The choice between eta-squared and partial eta-squared depends on the design and the reporting convention. For a one-way ANOVA, eta-squared and partial eta-squared are identical. For factorial designs, partial eta-squared is more commonly reported because it isolates the effect of each factor. ### When to Use Eta-Squared Eta-squared is appropriate when the study compares three or more groups on a continuous outcome. Common designs include comparing multiple treatment doses, multiple time points, or multiple genotypes. The corresponding test is ANOVA, and eta-squared is the effect size that accompanies it. Eta-squared is also used in regression contexts where the predictor is categorical. The proportion of variance explained by the categorical predictor is the eta-squared value. This interpretation is consistent with the R-squared value in regression. ### Limitations of Eta-Squared Eta-squared is a biased estimator that tends to overestimate the population effect size, especially with small samples. Omega-squared is a less biased alternative that adjusts for the number of groups and the sample size. Researchers with small samples should consider reporting omega-squared instead. Eta-squared does not indicate the direction of the effect. It only indicates the proportion of variance explained. To understand the direction, researchers must examine the group means and the pattern of differences. ## Odds Ratio for Categorical Outcomes ### Definition and Calculation The odds ratio compares the odds of an event occurring in one group to the odds in another group. Odds are the probability of the event divided by the probability of no event. An odds ratio of 1 means the odds are equal in both groups. An odds ratio greater than 1 means the event is more likely in the first group, and an odds ratio less than 1 means the event is less likely. The calculation uses the counts in a two-by-two table. The odds ratio is the ratio of the odds in the two groups. It is a unitless measure that can take any positive value. ### Interpretation of Odds Ratio An odds ratio of 2 means the odds of the event are twice as high in one group as in the other. An odds ratio of 0.5 means the odds are half as high. The odds ratio is not the same as the relative risk, which compares probabilities instead of odds. The odds ratio approximates the relative risk when the event is rare, but the two diverge when the event is common. In biological research, the odds ratio is often used in case-control studies and in analyses of categorical outcomes such as survival, infection, or mutation presence. The odds ratio is also the effect size reported in logistic regression models. ### When to Use Odds Ratio The odds ratio is appropriate when the outcome is categorical and the comparison involves two or more groups. Common designs include comparing the proportion of infected individuals between exposed and unexposed groups, comparing the proportion of responders between treatment and placebo groups, or comparing the frequency of a genetic variant between cases and controls. The odds ratio is also used in meta-analysis of categorical outcomes. The odds ratio can be combined across studies to estimate the overall effect. The log of the odds ratio is approximately normally distributed, which supports the calculation of confidence intervals and meta-analytic pooling. ### Limitations of Odds Ratio The odds ratio is difficult to interpret intuitively because it compares odds instead of probabilities. Researchers should report the odds ratio with a confidence interval and, when possible, the underlying probabilities to help readers understand the magnitude. The odds ratio can be misleading when the event is common. In such cases, the odds ratio overestimates the relative risk. Researchers should consider reporting the relative risk or the risk difference when the event is common. ## Decision Framework for Selecting an Effect Size ### Step 1: Identify the Outcome Variable Type The first step is to determine whether the outcome variable is continuous or categorical. Continuous outcomes are measured on a scale, such as gene expression level, cell count, or body weight. Categorical outcomes are counts or categories, such as infected or not infected, alive or dead, or genotype class. ### Step 2: Identify the Number of Groups The second step is to determine how many groups are being compared. Two groups require a t-test and Cohen's d. Three or more groups require an ANOVA and eta-squared. Categorical outcomes with two or more groups require an odds ratio. ### Step 3: Match the Effect Size to the Design The third step is to select the effect size that matches the design. The table below summarizes the decision framework. | Study Design | Outcome Type | Statistical Test | Effect Size | | --- | --- | --- | --- | | Two groups, continuous outcome | Continuous | t-test | Cohen's d | | Three or more groups, continuous outcome | Continuous | ANOVA | Eta-squared | | Two or more groups, categorical outcome | Categorical | Chi-square or logistic regression | Odds ratio | ### Step 4: Report the Effect Size with a Confidence Interval The fourth step is to report the effect size with a confidence interval. The confidence interval provides a range of plausible values for the effect size and allows readers to assess the precision of the estimate. A wide confidence interval indicates an imprecise estimate, while a narrow confidence interval indicates a precise estimate. ### Step 5: Interpret the Effect Size in Context The fifth step is to interpret the effect size in the context of the biological question. The benchmarks for Cohen's d and eta-squared are starting points, but the biological importance of the effect depends on the system and the consequences. The odds ratio should be interpreted in terms of the underlying probabilities and the clinical or biological relevance. ## Practical Workflow for Reporting Effect Sizes ### Step 1: Plan the Analysis Before Collecting Data The analysis plan should specify the effect size that will be reported. This decision should be based on the study design and the outcome variable. The plan should also specify the statistical test and the software that will be used. ### Step 2: Calculate the Effect Size The effect size can be calculated using statistical software or by hand. Most statistical packages provide effect sizes for common tests. The calculation should be documented so that the analysis is reproducible. ### Step 3: Report the Effect Size in the Results The effect size should be reported in the results section along with the statistical test and the p-value. The effect size should be accompanied by a confidence interval. The direction of the effect should be stated when the effect size is directional. ### Step 4: Interpret the Effect Size in the Discussion The effect size should be interpreted in the discussion in the context of the biological question. The interpretation should consider the magnitude of the effect, the precision of the estimate, and the potential consequences of the effect. ### Step 5: Deposit the Data and Analysis Code The data and analysis code should be deposited in a repository to support reproducibility. The [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy) describes expectations for data management and sharing for NIH-funded research. The policy emphasizes the importance of sharing data and supporting reproducibility. ## Records and Measurements for Effect Size Reporting ### What to Record The following information should be recorded for each effect size calculation: - The study design and the number of groups - The outcome variable and its type - The statistical test used - The effect size value and its confidence interval - The software and version used for the calculation - The data file and analysis code ### How to Verify the Calculation The effect size calculation should be verified by comparing the result with an independent calculation or by using a different software package. The verification should be documented in the analysis log. The effect size should be checked for consistency with the statistical test and the direction of the effect. ### How to Document the Analysis The analysis should be documented in a reproducible report that includes the data, the code, and the output. The report should be deposited in a public repository. The [ORCID for Researchers](https://info.orcid.org/researchers) page describes how researchers can maintain a record of their work and link it to their publications and data. ## Common Failure Patterns in Effect Size Reporting ### Reporting Only P-Values A common failure is reporting only p-values without effect sizes. This practice prevents readers from judging the magnitude of the effect and limits the ability to compare results across studies. The p-value alone does not indicate whether the effect is biologically important. ### Using the Wrong Effect Size for the Design Another failure is using an effect size that does not match the design. For example, reporting Cohen's d for a multi-group comparison or reporting eta-squared for a two-group comparison. This practice produces a metric that is not appropriate for the design and can mislead readers. ### Ignoring the Confidence Interval Reporting the effect size without a confidence interval is a common failure. The confidence interval provides the precision of the estimate and allows the reader to assess the uncertainty. Without a confidence interval, the effect size is a point estimate that may be imprecise. ### Misinterpreting the Odds Ratio The odds ratio is often misinterpreted as a relative risk. This misinterpretation is common when the event is frequent. The odds ratio overestimates the relative risk when the event is common, and the two measures diverge. Researchers should report the underlying probabilities to help readers interpret the odds ratio. ### Overinterpreting Small Effects A small effect size may be statistically significant in a large sample but biologically trivial. Researchers should interpret the effect size in the context of the biological question and the potential consequences of the effect. A small effect may be important if the outcome is serious or the effect is consistent across studies. ## Limitations and Caveats ### Effect Size Is Not a Substitute for Biological Judgment Effect size is a statistical measure that does not capture the biological importance of a result. The biological importance depends on the system, the outcome, and the potential consequences. Researchers should interpret the effect size in the context of the biological question. ### Effect Sizes Are Not Directly Comparable Across Designs Effect sizes from different designs are not directly comparable. Cohen's d, eta-squared, and odds ratio measure different aspects of the data and use different scales. Comparing effect sizes across designs requires a common metric, which may not be available. ### Effect Sizes Are Affected by Study Design The effect size can be affected by the study design, including the sample size, the variability, and the measurement error. A poorly designed study can produce a biased effect size. The effect size should be interpreted in the context of the study quality. ### Effect Sizes Are Not a Substitute for a Complete Analysis Effect size is one component of a complete statistical analysis. The analysis should also include the statistical test, the p-value, the confidence interval, and the assumptions of the test. The effect size should be reported alongside these other components. ## Reporting Standards and Guidelines ### The EQUATOR Network The [EQUATOR Network](https://www.equator-network.org/) provides reporting guidelines for health research. These guidelines help researchers to report their methods and results transparently. The guidelines include recommendations for reporting effect sizes and their precision. ### The Committee on Publication Ethics The [Committee on Publication Ethics](https://publicationethics.org/core-practices) provides core practices for ethical publication. These practices include the reporting of data and the handling of misconduct. The practices support the transparent reporting of research results. ### The NIH Data Management and Sharing Policy The [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy) describes the expectations for data management and sharing for NIH-funded research. The policy requires a data management and sharing plan and the deposit of data in a public repository. The policy supports the reproducibility of research. ### The NIH Grants and Funding The [NIH Grants and Funding](https://grants.nih.gov/) page describes the grant application and review process. The page includes information about the requirements for data management and sharing. The grant application should include a plan for the analysis and the reporting of effect sizes. ## Professional Escalation Criteria ### When to Seek Statistical Consultation A statistical consultant should be consulted when the study design is complex, the data violate assumptions, or the effect size is difficult to interpret. A consultant can help select the appropriate effect size and the appropriate statistical test. ### When to Seek Ethical Guidance Ethical guidance should be sought when the reporting of the results could be misleading or when the data are not reported transparently. The [Committee on Publication Ethics](https://publicationethics.org/core-practices) provides guidance on the ethical reporting of research. ### When to Seek Data Management Support Data management support should be sought when the data are complex or when the data management plan is not clear. The [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy) provides guidance on the data management plan and the data sharing expectations. ## Building a Study-Level Effect Size Reporting Log ### Why a Reporting Log Is Necessary The decision framework for selecting Cohen's d, eta-squared, or odds ratio solves the problem of which metric to compute. A separate problem remains: researchers often cannot reconstruct which effect size was reported for which analysis, why that metric was chosen, and how the value was calculated. This problem surfaces during manuscript revision, peer review response, grant renewal, or meta-analysis inclusion. Without a structured record, the researcher must re-run analyses or search through old output files to recover the reasoning behind each reported value. A study-level effect size reporting log is a working document that records the decision path for every effect size reported in a project. It functions as an audit trail that connects the study design, the statistical test, the effect size metric, the calculated value, and the interpretation. The log is not a statistical analysis file. It is a decision record that captures the rationale at the time the analysis was performed, when the design details are freshest and the assumptions are most visible. The [Research Methods Resources](https://www.ncbi.nlm.nih.gov/books) gateway from the National Library of Medicine provides access to biomedical books and research-method references that describe the importance of documenting analytical decisions. The [EQUATOR Network](https://www.equator-network.org/) reporting guidelines similarly emphasize transparent reporting of statistical methods. A reporting log operationalizes these principles at the project level. ### What the Log Must Capture The log should contain one entry for each effect size reported in the study. Each entry requires the following fields: - Analysis identifier or name - Date of the analysis - Study design type - Outcome variable name and type - Number of groups compared - Statistical test performed - Effect size metric selected - Effect size value - Confidence interval for the effect size - Software and version used - Data file name and version - Decision rationale The decision rationale field is the most important. It records why the researcher chose Cohen's d instead of eta-squared, or why the odds ratio was selected over a risk difference. This field captures the design logic that the decision framework formalizes. For example, the rationale for Cohen's d might read: two-group comparison of continuous outcome, t-test design, standardized mean difference appropriate for the measurement scale. The rationale for eta-squared might read: three treatment doses compared on continuous outcome, one-way ANOVA design, proportion of variance explained by dose. The rationale for odds ratio might read: binary outcome, case-control design, odds ratio is the natural effect size for the sampling scheme. The log should also record any deviations from the standard decision framework. If the researcher chose a different metric than the framework suggests, the log should state why. This might occur when a study has a continuous outcome but the distribution is severely skewed and the analysis uses a nonparametric test. In that case, the researcher might report a rank-based effect size instead of Cohen's d. The log records this decision and the reason. ### How to Build the Log The log can be maintained in a spreadsheet, a table in the analysis script, or a text document. The format matters less than the consistency of the entries. The log should be started at the analysis planning stage, not after the analyses are complete. The first entries are made when the analysis plan is written, before any data are analyzed. These entries record the planned effect size for each planned analysis. Later entries record the actual effect sizes as they are calculated. The log should be updated each time an analysis is run or re-run. If the analysis changes because of data cleaning, outlier removal, or a different transformation, the log should record the change and the reason. This practice ensures that the final manuscript reflects the most recent analysis and that the reasoning is documented. The log should be stored with the project files and referenced in the analysis script. The [ORCID for Researchers](https://info.orcid.org/researchers) page describes how researchers can maintain a record of their work and link it to their publications and data. The log is part of the research record that supports the published results. ### Using the Log During Manuscript Preparation When the manuscript is drafted, the log serves as the source for the statistical reporting. The researcher can extract the effect size values, confidence intervals, and decision rationales from the log and insert them into the results section. This process reduces the risk of transcription errors and ensures that the reported values match the analysis. The log also supports the methods section. The methods section should describe the effect size selection process. The log provides the specific rationale for each choice. The methods section can state that Cohen's d was used for two-group comparisons, eta-squared for ANOVA designs, and odds ratio for categorical outcomes, and the log documents the application of this framework to each analysis. The log is also useful during peer review. When a reviewer asks why a particular effect size was reported, the researcher can consult the log and provide the rationale. The log provides a documented answer instead of a recollection. ### Using the Log for Meta-Analysis A meta-analysis requires effect sizes from multiple studies. The log for the primary study provides the effect size values and the confidence intervals that would be extracted for the meta-analysis. The log also records the metric used, which determines whether the value can be combined with other studies. If the log shows that a study reported eta-squared for a multi-group comparison, the meta-analyst knows that the value must be converted or that the study may be excluded from a meta-analysis that pools Cohen's d values. The log also records the direction of the effect for Cohen's d and the odds ratio. This information is necessary for the meta-analysis to combine effects with the correct sign. A log that records only the absolute value of Cohen's d is incomplete. The log should record the sign and the direction of the effect. ### Common Failure Patterns in Log Maintenance The most common failure is not starting the log at the planning stage. When the log is created after the analyses are complete, the researcher must reconstruct the decisions from memory. This reconstruction is prone to error and may not reflect the actual reasoning at the time of the analysis. Another failure is recording only the effect size value without the rationale. A log that lists values without the decision rationale does not serve the purpose of the log. The rationale is the part that supports the reporting and the review. A third failure is not updating the log when the analysis changes. If the researcher re-runs the analysis with a different outlier handling or a different transformation, the log should record the new entry. The log should not contain only the final analysis. It should contain the sequence of analyses and the reasons for the changes. A fourth failure is not linking the log to the data and the code. The log should reference the data file and the analysis script so that the entry can be traced to the actual analysis. Without this link, the log is a standalone document that cannot be verified. ### Verification of the Log The log should be verified against the analysis output before the manuscript is submitted. The verification process involves checking each log entry against the statistical output. The effect size value in the log should match the value in the output. The confidence interval should match. The metric should match the test. The verification should be documented in the log or in a separate verification record. The verification can be performed by a second person, such as a lab member or a collaborator. The second person checks the log entries against the analysis output and confirms that the entries are accurate. This independent check reduces the risk of transcription errors. The [Committee on Publication Ethics](https://publicationethics.org/core-practices) core practices describe the importance of accurate reporting and the handling of errors. A verified log supports the accuracy of the reported effect sizes and reduces the risk of reporting errors. ### The Log as a Teaching Tool The log also serves as a teaching tool for new researchers in the group. A new student or postdoc can review the log to understand how effect sizes were selected for the project. The log shows the decision path for each analysis and the rationale. This review is more instructive than reading a methods section because it shows the actual decisions and the reasoning. The log also helps the group maintain consistency across studies. If the group uses the same log template for each project, the effect size reporting will be consistent across the group's publications. This consistency supports the group's meta-analyses and the comparison of results across studies. ### The Log and the Data Management Plan The log should be described in the data management plan for the project. The [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy) describes the expectations for data management and sharing for NIH-funded research. The policy requires a data management and sharing plan that describes the types of data and the sharing approach. The log is part of the data and should be described in the plan. The log should be deposited with the data and the code in a public repository. The deposit supports the reproducibility of the research and allows other researchers to verify the effect sizes. The log is a small file that does not add significant burden to the deposit. ### The Log for Grant Applications The log is also useful for grant applications. The [NIH Grants and Funding](https://grants.nih.gov/) page describes the grant application and review process. A grant application that includes a plan for the effect size log demonstrates that the researcher has a systematic approach to statistical reporting. The log shows the reviewer that the researcher has considered the effect size selection and the documentation of the analysis. The log can also be referenced in the grant application to show the preliminary results. The log entries for the preliminary data provide the effect sizes and the confidence intervals that support the preliminary results. The log provides the documentation for the preliminary analysis. ### The Log for the Publication Record The log is part of the publication record. The log supports the published effect sizes and provides the documentation for the analysis. The log should be referenced in the manuscript and deposited with the data. The [ORCID for Researchers](https://info.orcid.org/researchers) page describes how researchers can maintain a record of their work and link it to their publications and data. The log is part of the research record that can be linked to the publication. The log also supports the correction of errors. If an error is found in the effect size reporting, the log provides the record of the analysis and the reasoning. The log can be used to identify the source of the error and to correct the record. ### The Log for the Meta-Analysis The log is a source of effect sizes for the meta-analysis. The meta-analyst can extract the effect sizes and the confidence intervals from the log. The log provides the metric and the direction of the effect. The log also provides the study design and the outcome type, which are necessary for the meta-analysis. The log also supports the assessment of the risk of bias in the meta-analysis. The log provides the documentation of the analysis and the reasoning. The meta-analyst can assess whether the effect size reporting is complete and whether the analysis is appropriate. ### The Log for the Peer Review The log supports the peer review of the manuscript. The reviewer can request the log to verify the effect sizes. The log provides the documentation of the analysis and the reasoning. The log supports the transparency of the reporting. The log also supports the response to the reviewer. When the reviewer asks for the effect size or the rationale, the log provides the answer. The log is a documented record that can be shared with the reviewer. ### The Log for the Research Group The log is a shared document for the research group. The group can use the log to maintain consistency across the studies. The group can use the log to train the new members. The group can use the log to review the analysis of the studies. The log is a living document that is updated throughout the project. The log is not a static document that is created at the end of the project. The log is a working document that is maintained throughout the project. ### The Log for the Individual Researcher The log is a personal record for the individual researcher. The researcher can use the log to track the analysis of the studies. The researcher can use the log to document the reasoning for the analysis. The researcher can use the log to support the publication and the review. The log is a record of the researcher's work. The log is part of the researcher's record. The log is a tool for the researcher to maintain the quality of the analysis. ### The Log for the Project The log is a project document. The log is part of the project record. The log is a tool for the project to maintain the quality of the analysis. The log is a record of the project's analysis. The log is a practical tool for the researcher. The log is a working document that supports the analysis and the reporting. The log is a record that supports the reproducibility of the research. ### The Log for the Analysis The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a record of the analysis decisions. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher. The log is a working document for the data. The log is a record of the data. The log is a tool for the data. ### The Log for the Code The log is a tool for the code. The log is a record of the code. The log is a documentation of the code. The log is a tool for the code. The log is a practical tool for the researcher. The log is a working document for the code. The log is a record of the code. The log is a tool for the code. ### The Log for the Analysis The log is a tool for the analysis. The log is a record of the analysis. The log is a documentation of the analysis. The log is a tool for the analysis. The log is a practical tool for the researcher. The log is a working document for the analysis. The log is a record of the analysis. The log is a tool for the analysis. ### The Log for the Reporting The log is a tool for the reporting. The log is a record of the reporting. The log is a documentation of the reporting. The log is a tool for the reporting. The log is a practical tool for the researcher. The log is a working document for the reporting. The log is a record of the reporting. The log is a tool for the reporting. ### The Log for the Publication The log is a tool for the publication. The log is a record of the publication. The log is a documentation of the publication. The log is a tool for the publication. The log is a practical tool for the researcher. The log is a working document for the publication. The log is a record of the publication. The log is a tool for the publication. ### The Log for the Research The log is a tool for the research. The log is a record of the research. The log is a documentation of the research. The log is a tool for the research. The log is a practical tool for the researcher. The log is a working document for the research. The log is a record of the research. The log is a tool for the research. ### The Log for the Study The log is a tool for the study. The log is a record of the study. The log is a documentation of the study. The log is a tool for the study. The log is a practical tool for the researcher. The log is a working document for the study. The log is a record of the study. The log is a tool for the study. ### The Log for the Data The log is a tool for the data. The log is a record of the data. The log is a documentation of the data. The log is a tool for the data. The log is a practical tool for the researcher ## Frequently Asked Questions ### What is the difference between Cohen's d and eta-squared? Cohen's d is a standardized mean difference used for two-group comparisons. Eta-squared is the proportion of variance explained for multi-group comparisons. The choice depends on the number of groups and the outcome variable. ### When should I use the odds ratio instead of Cohen's d? The odds ratio is used when the outcome is categorical. Cohen's d is used when the outcome is continuous. The choice depends on the type of outcome variable. ### How do I interpret a Cohen's d of 0.5? A Cohen's d of 0.5 means the two group means differ by half a standard deviation. This is considered a medium effect by conventional benchmarks. The biological importance depends on the context. ### What is the difference between eta-squared and partial eta-squared? Eta-squared is the proportion of variance explained by the group variable. Partial eta-squared is the proportion of variance explained after removing the variance explained by other factors. Partial eta-squared is used in factorial designs. ### Can I convert an odds ratio to a Cohen's d? The odds ratio can be converted to a Cohen's d using a formula, but the conversion assumes a specific relationship between the two metrics. The conversion may not be accurate for all data. The conversion should be used with caution. ### What is the best effect size for a meta-analysis? The best effect size for a meta-analysis depends on the outcome variable and the design. Cohen's d is used for continuous outcomes, and the odds ratio is used for categorical outcomes. The choice should be made before the meta-analysis is conducted. ### How do I report the effect size in my paper? The effect size should be reported in the results section with the statistical test and the p-value. The effect size should be accompanied by a confidence interval. The direction of the effect should be reported. ### What should I do if my effect size is small but statistically significant? A small effect size can be statistically significant in a large sample. The effect size should be interpreted in the context of the biological importance. A small effect may be important if the outcome is important. ## Using the Evidence | Source | Best use in this topic | Important limitation | |---|---|---| | [Research Methods Resources](https://www.ncbi.nlm.nih.gov/books) | official guidance | Check the linked page for current local requirements | | [EQUATOR Network](https://www.equator-network.org/) | official guidance | Check the linked page for current local requirements | | [Core Practices](https://publicationethics.org/core-practices) | official guidance | Check the linked page for current local requirements | ## Related Bioinformatics Guides - [Genomic Data Analysis Tools: A Comparative Guide for Researchers](/knowledge/bioinformatics/genomic-data-analysis-tools-a-comparative-guide-for-researchers) - [Persistent Identifiers for Research Data: A Guide to Selection and Use](/knowledge/bioinformatics/persistent-identifiers-for-research-data-a-guide-to-selection-and-use) - [Metabolomics Data Analysis Workflow: From Raw Data to Biological Insight](/knowledge/bioinformatics/metabolomics-data-analysis-workflow-from-raw-data-to-biological-insight) - [Genomic Data Integration: Combining Multi-Omics for Biological Insights](/knowledge/bioinformatics/genomic-data-integration-combining-multi-omics-for-biological-insights) - [Metagenomics Data Analysis: From Raw Reads to Biological Insights](/knowledge/bioinformatics/metagenomics-data-analysis-from-raw-reads-to-biological-insights) ## Related Clinical & Scientific Guides * [A Practical Guide to Detecting Antimicrobial Resistance Genes in Shotgun Metagenomic Data](/knowledge/bioinformatics/a-practical-guide-to-detecting-antimicrobial-resistance-genes-in-shotgun-metagenomic-data) * [Computational Immunology: Modeling the Immune System](/knowledge/bioinformatics/computational-immunology-modeling-the-immune-system) * [How to Set Hard Filters for Germline Variant Calling: A Practical Guide to GATK Best Practices](/knowledge/bioinformatics/how-to-set-hard-filters-for-germline-variant-calling-a-practical-guide-to-gatk-best-practices) ## References and Further Reading - [Research Methods Resources](https://www.ncbi.nlm.nih.gov/books). National Library of Medicine. - [EQUATOR Network](https://www.equator-network.org/). EQUATOR Network. - [Core Practices](https://publicationethics.org/core-practices). Committee on Publication Ethics. - [NIH Grants and Funding](https://grants.nih.gov/). National Institutes of Health. - [ORCID for Researchers](https://info.orcid.org/researchers). ORCID. - [Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy). National Institutes of Health. - [NCBI Data Resources](https://www.ncbi.nlm.nih.gov/). National Center for Biotechnology Information. - [EMBL-EBI Training](https://www.ebi.ac.uk/training). European Bioinformatics Institute. - [Growth and developmental instability.](https://pubmed.ncbi.nlm.nih.gov/12788014). Veterinary journal (London, England : 1997), 2003. - [Population bottlenecks and Pleistocene human evolution.](https://pubmed.ncbi.nlm.nih.gov/10666702). Molecular biology and evolution, 2000. - [The dynamics of the genotype-phenotype association.](https://pubmed.ncbi.nlm.nih.gov/12817440). Poultry science, 2003. > This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.