### INTRODUCTION

*in vivo*trials. In these studies, statistical methods are essential, and authors should strive to employ an appropriate experimental design and statistical analyses to provide the reader with scientifically relevant and valid knowledge.

### EXPERIMENTAL DESIGN

### The experimental unit and the number of replicates

### Consideration for known variations

### Randomization

### STATISTICAL ANALYSIS

### Statistical models

### Statistical methods

### Comparing the means of interest

### Statistical software packages and statistical procedures

### REPORTING

### SUMMARY OF RECOMMENDATIONS

Provide details of experimental design and statistical methods in the Materials and Methods section.

Define the experimental unit and report the number of replicates. Replicates are to be independent observations and experimental units must be independent of each other.

Conduct power tests and provide their results to justify the experiment was appropriately designed.

Use blocks or covariates whenever applicable to reduce unexplained experimental errors.

Describe the procedure used to minimize possible carryover effects and to show carryover effects are not significant when using a crossover design.

Ensure the implementation of randomization when sampling experimental units and allocating treatments to experimental units.

Describe the statistical models used for the statistical analysis as equations, as well as in words.

Use appropriate statistical methods depending on the data type of observations, research questions to be answered, and the statistical model.

Test if the observations are normally distributed around their mean. If not or the sample size is not large enough, use non-parametric analyses instead; otherwise, use parametric methods.

State the level of statistical significance in the statistical analysis section.

Conduct post hoc analysis on the basis that there is a significant difference among the treatments and use appropriate methods according to the structure of the treatments.

Perform pair-wise comparisons (e.g., Duncan’s multiple range test) only when there is no structure among a series of treatments.

Report which software and procedures are used for the statistical analysis.

Use appropriate statistical methods and procedures if observations are repeatedly measured or random effects are expected.

Present both probability value (p-value) and pooled SEM as error measures. The standard deviation can only be used for descriptive statistics.

Report outliers or missing data in the Materials and Methods section or the Results section where it is more appropriate.