Nutritional surveys based on memory bear little resemblance to what was actually consumed. From ACSH, Memory-based recall isn’t a solid platform for dietary recommendations
Nutritional studies based on surveys and memory have trouble being reproduced because memory and actual may bear little resemblance to each other. People don’t remember what they ate and, I’ll add, do not measure and poorly estimate quantities. So, unless you are keeping a diary with weights and measures, just how good would a survey asking how much and what you ate this year?
I’d add that studies based on surveys on historical use of chemicals fall into a similar category. Several years ago I responded a survey about agricultural chemicals I might have been exposed to in the 1960’s. Just how valid do you suppose the results of that survey were?