Experimental design, data collection and statistical analysis
The Markey Cancer Center’s (MCC) Biostatistics Shared Resource Facility (BSRF) collaborated with the MCC Small Animal Imaging Facility to provide guidance regarding experimental design, data collection and statistical analysis of xenograft, transgenic and other mouse models of cancer.
The goal is to ensure that studies are designed properly in order to adequately address research hypotheses and collect and store high-quality data.
Basic elements for designing experiments:
- Research objectives and translation into testable statistical hypothesis.
- Primary study endpoint. Secondary and exploratory outcomes.
- Number of groups, choice of control groups
- Animal model and features/characteristics of animal model
- Study Design: Two or multiple parallel groups, Factorial design, Longitudinal study, Paired design
- Sample size (# of animals per group):
- Effect size: Hypothesized amount of difference between groups and amount of variation in study outcome
- Statistical power desired (at least 80%) and significance level (usually 5%)
- Sample size method consistent with study design and analysis plans
- ID/Coding of Animals
Data collection (LabKey):
We recommend that data collected in the MCCSAIF use the database system developed through LabKey. This will facilitate accurate, clean and complete data collection and data entry for timely statistical analysis.
Please refer to the LabKey link in this website for use of this database system.
Overview video of “Experiment design, data collection and statistical analysis”