Validation of dietary assessment methods relies on several complementary approaches that together determine whether a tool measures intake accurately, consistently, and for the intended purpose.
Face and content validity
This is the foundation of validation techniques, where experts assess whether the method appears appropriate and covers all relevant aspects of diet reporting.
Construct validity
Construct validity asks whether the tool behaves as expected when compared with related measures or known patterns.
- Comparison with related behaviours — For example, people with higher BMI or known high-energy diets should report higher intakes.
- Testing expected associations — If the tool is valid, nutrient intakes should correlate with biomarkers or health indicators in predictable ways.
Construct validity helps determine whether the tool is measuring the underlying concept of “dietary intake” rather than something unrelated.
Criterion validity (the gold standard for accuracy)
Criterion validity evaluates how close the tool’s results are to a trusted reference method. This is the most important category for assessing accuracy.
- Comparison with interviewer-led 24‑hour recalls or food diaries — A common approach for self-report tools.
- Comparison with objective biomarkers — Such as doubly labelled water for energy intake, urinary nitrogen for protein, or recovery biomarkers for potassium and sodium. These are considered the most rigorous tests because they bypass reporting bias.
- Comparison with weighed food records — Useful for portion-size accuracy and detailed intake patterns.
Reliability (consistency over time)
- Test–retest reliability — Participants complete the tool twice under similar conditions; consistent results indicate reliability.
- Internal consistency — Ensures different parts of the tool (e.g., portion-size questions) work together coherently.
- Day-to-day stability — Particularly important for 24‑hour recalls, where multiple administrations allow assessment of within-person variation.
Recommended approach for validation of Intake24 is a comparison with the existing dietary assessment method in the population e.g. food diaries or interviewer-led 24hour recalls. This is a cost-effective method of validation and tests the online tool against the current preferred assessment method.
- Relative validation against interviewer‑led multiple‑pass 24‑hour recalls A large Newcastle University study with 180 participants aged 11–24 compared Intake24 directly with traditional interviewer-led recalls conducted on the same day, repeated four times over a month. This design assessed whether the automated system could match the detail and quality of trained interviewers. Results showed that mean intakes of all macronutrients and micronutrients (except non-milk extrinsic sugars) were within 4% of the interviewer-led recall. Comparison of INTAKE24 (an Online 24-h Dietary Recall Tool) with Interviewer-Led 24-h Recall in 11–24 Year-Old | MDPI
Validations using biomarkers are recommended as they provide an objective comparison of intake. However, these studies are costly to run, and there is a limited availability of validated biomarkers. This approach does not indicate the types of foods being omitted from the Intake24 recall, which is important for system development. A validation comprising of both self-reported and biomarkers would strengthen the approach.
- Validation against doubly labelled water (DLW) A validation of Intake24 using DLW in adults, found that the system underestimated energy by approximately 25% relative to DLW – this level of underreporting is similar to other self-reported methods. Repeated Intake24 recalls showed high internal consistency and day-to-day variation was captured appropriately. Validity and reliability of an online self-report 24-h dietary recall method (Intake24): a doubly labelled water study and repeated-measures analysis | Journal of Nutritional Science | Cambridge Core
Controlled feeding studies are one of the most rigorous ways to validate dietary assessment methods because they create a setting where true intake is known with high precision. However this approach is costly to implement and has high participant burden.
- A controlled feeding study in 152 participants, assessed four technology assisted dietary assessment methods- Automated Self-Administered Dietary Assessment Tool-Australia (ASA24), Intake24-Australia; mobile Food Record-Trained Analyst (mFR-TA); and Image-Assisted Interviewer-Administered 24-hour recall (IA-24HR). Under controlled conditions, Intake24, ASA24 and mFR-TA estimated average energy and nutrient intakes with reasonable validity, but intake distributions were estimated accurately by Intake24 only (energy and protein). Accuracy of energy and nutrient intake estimation versus observed intake using 4 technology-assisted dietary assessment methods: a randomized crossover feeding study - ScienceDirect
- A study by Sharon Kirkpatrick, Amy Subar and colleagues, evaluated the criterion validity of ASA24, using a controlled feeding design in which true intake was known. Eighty‑one adults consumed three meals from a buffet where all foods and beverages were unobtrusively weighed before and after self‑service to establish precise intake. Participants then completed both an ASA24 self-administered 24‑hour recall and an interviewer-administered AMPM recall. The study found that ASA24 performed comparably to interviewer-led recalls, with both methods showing the expected under‑reporting relative to true intake. Agreement in reported foods and nutrient estimates was high, demonstrating that ASA24 can capture dietary intake with accuracy similar to trained interviewers. This provided strong evidence that a fully automated, web-based recall can serve as a valid alternative to interviewer-administered methods in research settings. Performance of the Automated Self-Administered 24-hour Recall relative to a measure of true intakes and to an interviewer-administered 24-h recall - ScienceDirect