I notice biostatisticians are not told their thesis is ‘just’ data analysis and doesn’t mean anything if they didn’t actually personally recruit patients into a clinical trial, etc. Understanding the nature and complexity of the data one is analysing is important, personally generating it is not.

Regrettably the data side of science is not as valued by most instituitions as the wet lab stuff. It's a sector wide problem and partly why there are endemic issues with recruitment and retention.