Here's how you can make your feedback on a statistical research proposal clear and actionable.

Before diving into the specifics of the proposal, make sure you fully understand the research objectives. Ask yourself what the researcher aims to discover or demonstrate with their study. Understanding the objectives allows your feedback to be focused and relevant. If the objectives are unclear, suggest ways to refine them. Remember, a well-defined objective is the compass that guides the entire research process, ensuring all statistical methods and data collection strategies align with the intended outcomes.

Often I find researchers (including myself), are not always practical when designing their study method (e.g., fail to consider timelines, publishing requirements, funds available, budget, staffing,and participation). One way to help circumvent this is to design and go over the standardized operating procedures (SOP) in advance with the project manager. In addition, consulting with a statistician can help ensure that the methods (including proposed number of participants), goals and the statistics are all aligned prior to grant or IRB submission.

One helpful way to ensure data collection is done using the right tools and methods is to examine the variables of study and goals closely in order not to miss any important details. Revisit them several times before data collection to avoid a costly mistake.

It is very important to give unbiased constructive feedback at this stage because the more the weaknesses the greater the error margins and hence rendering results useless and cannot be generalized to the population.

One helpful way is to encourage the researcher to revisit their previous hypothesis and research aims and making sure the way there results are discussed reflects on that

An example of the importance of giving feedback on research methods is when a postgraduate student only asked for feedback after collecting the data only to find that very important research questions were not included in the questionnaire and the sample size was not correct. That lead to her loosing a lot of time, money and effort.