Skip to main content

Experimental Design

When developing an EMSL user project proposal, reach out to EMSL staff to learn about pertinent considerations for the statistical design of your experiment.

By checking a box in the letter of intent submission form, you will indicate if you are interested in discussing statistical experimental design and statistical analysis methods, which can be incorporated into your written proposal. By identifying this potential need in the development phase, project needs can be identified early to efficiently plan a statistical design that addresses your scientific question of interest.

Experimental design can be considered from various perspectives, and the check box on the letter of intent submission form is focused primarily on the statistical perspective.

Domain science:

  • Question of interest
  • Scientific impact
  • Required resources, materials, and instruments

Statistical:

  • Data types: continuous, binary, count, etc.
  • Correlation structure: time course, spatial, etc.
  • Statistical power
  • Statistical analysis method

Reviewer:

  • Understanding the structure of the experiment and the connection to the broader scientific field
  • Proposed work reasonably addresses the scientific question
  • Impact

Research application

The Data Transformations Integrated Research Platform can work with EMSL users on statistical design of experiments, supporting research across all Integrated Research Platforms.

Available resources

Batch correction: This resource is available externally and utilized by EMSL staff to remove batch effects (that cause nonbiological differences between samples from different batches) from data:

Randomization: EMSL staff can help generate random ordering for sample preparation and instrument runs to minimize systematic bias and account for extreme variability to ensure unbiased downstream analyses.

Power analyses: EMSL staff can help calculate sample sizes needed to achieve the desired power, which is the ability to statistically detect a difference when the difference truly exists.

Tips for success

  • To maximize effectiveness and efficiency, engage with EMSL staff at the proposal development stage if you have questions about or would like to discuss statistical design of experiments or statistical analysis methods.
  • Statistical design of experiments at the proposal development stage helps assure reviewers that the experiment and resulting data will be able to address the scientific question of interest, makes project kick-off more efficient, and enables the project team to hit the ground running.
  • If sample numbers require multiple prep or run batches, it is helpful to identify this early in the project lifetime, as some batch correction methods require spiked in internal standards or that quality control samples be run periodically throughout each batch of samples. While the planning for batch effect correction is best done at the proposal development stage, it is actually performed during the project lifetime.

 

Contacts