Frequent food questionnaires (FFQs) are often given to large groups of people to obtain information on their dietary habits. Study participants are asked about the frequency with which they consume certain goods. Another method to obtain information on foods consumed is a food diary. People are asked to record every type of food and amount consumed for a few days. Food diaries are more difficult to obtain and response rates are lower than for FFQs. A study was conducted to see how well FFQs predict food consumed based on food diaries. A statistical package calculates confidence intervals for the mean number of drinks consumed per food diary based on answers from the FFQ, as well as prediction intervals. Prediction intervals are always:

wider than mean intervals.

narrower than mean intervals if the predicted number of drinks based on the FFQ is smaller than the true number consumed based on food diary and wider otherwise.

narrower than mean intervals.

equal to mean intervals.