Comparison of an Adapted Food Security Module for College Students with the United States Department of Agriculture Adult Food Security Module

FAQ: Measuring Food Insecurity Among College Students

Why is it important to accurately measure food insecurity among college students?

Accurately measuring food insecurity (FI) among college students is crucial because existing estimates vary widely (from 10% to 75%), suggesting significant issues with current measurement methods. Despite these inconsistent figures, research consistently demonstrates that FI negatively impacts students' well-being. This includes poor diet quality, diminished academic performance, and adverse mental and physical health outcomes, such as a higher risk of depression and chronic diseases. Without precise and consistent data, government organizations and higher education institutions are hesitant to allocate sufficient resources to address this widespread problem, leaving many students vulnerable.

What are the main challenges with using the standard USDA Adult Food Security Survey Module (AFSSM) for college students?

Several challenges arise when using the standard USDA 10-item Adult Food Security Survey Module (AFSSM) for college students:

• Confusing Terminology: Students often struggle with terms like "household" and "balanced meal." Defining "household" is particularly complex for students, as it can refer to various living arrangements (dorms, apartments with roommates, family homes), and these situations often change. The AFSSM's classification of an entire household as food insecure if one respondent is, can lead to misclassification if roommates don't pool resources.
• Financial vs. Non-Financial Factors: Questions that stipulate experiences must be due to "not enough money" are problematic. Students have diverse financial resources (parental support, meal plans, scholarships, wages) and often experience food access issues due to non-financial reasons like limited transportation, inadequate storage, or dining hall hours, which the standard module may not capture. Some students even use "convenience" or "time" as cover-ups for financial constraints.
• Reference Timeframe and Survey Choice: The timeframe referenced (e.g., 12-month vs. 30-day) significantly impacts prevalence rates, with shorter timeframes often showing higher FI. The choice between the 6-item, 10-item, or 18-item modules, and the use of a 2-item screener, also contributes to varied prevalence estimates.
• Violation of Theoretical Item Order: The AFSSM items are designed to progress from least to most severe. However, for college students, items like "balanced meal" (intended to be less severe) are often affirmed more frequently, and "hunger" items (intended to be more severe) are affirmed more than expected, indicating that students interpret the questions differently than the general population.
• Sampling Bias: Most studies use non-representative samples, oversampling certain demographics (females, White and Asian students, younger students) or focusing on single universities, which can lead to biased prevalence estimates.

How did the researchers adapt the AFSSM for college students?

The researchers adapted the AFSSM for college students through a mixed-methods study involving three phases:

• Individual Cognitive Interviews: Researchers conducted interviews with 20 students to understand their interpretation of survey items. This phase identified confusion around "household," "balanced meal," and differentiating between frugality and food insecurity. As a result, nine items were revised, and one was removed. Changes included clarifying the timeframe to "30 days (1 month)," simplifying phrases like "run short of money" to "run low on money," and defining "balanced meal."
• Group Cognitive Interviews: Thirteen group interviews with 42 students were conducted to gather feedback on the initial revisions. Students provided further suggestions, particularly emphasizing non-financial factors impacting FI. Seven items were changed based on this feedback, including bolding phrases related to financial barriers to emphasize the monetary aspect, and clarifying ambiguous phrases such as "food did not last." The "eat less" item was merged with "cut meals" due to redundancy.
• Survey Study: The adapted survey module was then tested on a large sample (n=2897) of students. This allowed researchers to compare the psychometric validity and prevalence rates between the official and adapted versions. Adjustments were also made to the "household" and "annual income" questions, removing "household" and introducing a standardized template for income estimation due to students' multiple income streams.

What specific wording changes were made to the adapted AFSSM questions?

Key wording changes in the adapted AFSSM questions include:

• Timeframe: "In the last 30 d" was changed to "In the last 30 d (1 mo)" for all food security items for clarity.
• SQ1 (First Screener): "run short of money and try to make your food or your food money go further" became "run low on money to buy food" to simplify and avoid a double-barreled question.
• HH1 (Second Screener): "food eaten in your household" became "foods you ate" to address confusion around "household." Answer options were modified (e.g., "Enough, but not always the kind we want" had "always" removed) to avoid over-selection.
• HH2 (Worry): "worried whether" became "worried that," and "I often worried" was added to response options. An asterisk note was added to consider "food from all sources, including purchased food, foods from campus dining plans, food assistance programs, etc." The word "stressed" was added alongside "worried."
• HH3 (Food not lasting): "just didn't last" was changed to "ran out of food" to mimic HH2's phrasing.
• HH4 (Balanced Meal): A clear definition for "balanced meal" was added (grain, fruit/vegetable, protein food) with a prompt: "Please answer according to your ability to buy balanced meals and not based on your eating preferences."
• AD1 (Cut Meals): "eat less" was integrated into this question, reading "did you ever eat less (cut the size of your meals, skipped meals, etc.) because there wasn't enough money for food?" to combine two previously separate items.
• AD1a (Cut Meals, frequency) & AD5a (Not eating for the whole day, frequency): Response options were changed from requiring a specific number of days to ranges (e.g., 1-2 days, 3-7 days, etc.) to make answering easier and more accurate.
• AD4 (Lose Weight): Modified to include "gain or lose weight" and later to "noticed physical changes (i.e., changes in weight, clothes fitting differently, etc.)" to capture broader experiences beyond just weight loss.
• AD5 (Not eating for the whole day): Clarified to mean "not eat or drink anything (other than water) for a whole day."
• General: Phrases related to financial barriers (e.g., "because there wasn't enough money for food") were bolded in all food security items to emphasize the financial aspect.

What was the impact of the adapted AFSSM on food insecurity prevalence estimates among college students?

The study found a notable difference in food insecurity prevalence estimates. Using the official AFSSM, 35.4% of students were classified as food insecure. In contrast, using the adapted AFSSM, the prevalence was lower at 28.4% (P < 0.05). This suggests that the adapted module may lead to more precise or potentially lower estimates of food insecurity compared to the official version. The study concluded that while both versions provide psychometrically valid measures, adjustments might be needed before directly comparing FI measures for college students with those for the general population, due to the different ways questions like "food not lasting" and "balanced meals" captured FI prevalence.

Did the adapted AFSSM improve how students understood the concept of "household" and "annual income"?

Despite the adaptations, students still found defining "household" and estimating "annual income" challenging. The term "household" was removed from all survey items, and a new definition was created to prompt participants to count adults with whom they share all living expenses and financially dependent children. While this was intended to simplify, students still struggled with the concept. Similarly, adaptations were made to income measures, including a standardized template for estimating annual income from various sources (employment, loans, scholarships, family funds, SNAP). However, participants still felt their estimates were likely inaccurate, particularly for hourly jobs with inconsistent pay or random financial support from friends and family. The concerns about accurately measuring student income were not fully resolved.

What were the key findings regarding the "balanced meal" question (HH4) and its performance in the adapted module?

The "balanced meal" question (HH4) continued to be the most frequently affirmed item in both the official and adapted versions of the survey, consistent with prior research. Although the adapted version included a definition of a balanced meal and a prompt to answer based on financial ability rather than eating preferences, it did not significantly change the item's performance in the Rasch model. Regardless of the wording, this question captured a lower level of food insecurity severity than commonly found in general population food security measures. This suggests that even with clarifications, the item may not effectively differentiate between the "quality" of the diet for food-secure versus food-insecure students, implying that factors beyond finances (like personal preferences or access issues) might heavily influence student responses, making it less useful for measuring financial-based food insecurity in this population.

What are the study's limitations and future recommendations for measuring food insecurity in college students?

Limitations:

• Small Sample Sizes: Individual and group cognitive interviews had smaller sample sizes than desired, potentially limiting the diversity of perspectives captured, especially regarding international students or those living in dorms during the individual interviews.
• Institution Type: Data was collected only from four-year institutions, meaning the unique experiences of students at two-year institutions, who may have different FI experiences, were not captured.
• Continued Ambiguity: Despite efforts, some core issues like the interpretation of "household" and accurate income estimation remained problematic for students.

Future Recommendations:

• Substantive Changes: More significant changes to the AFSSM, potentially including new items or an adjusted scoring system (focusing on very low food security), might be needed to better distinguish students most in need.
• Concurrent Use: While the official AFSSM may misclassify some students, its continued use allows for comparisons with other studies and the general adult population. Therefore, other measures could be used in conjunction with the AFSSM.
• Holistic Approaches: Implement more holistic approaches that assess the full range of issues related to food access beyond just finances. This could involve testing complementary measures that evaluate resilience, food availability, utilization, stability, and nutrition security.
• Frequent Testing: Given that students' financial circumstances can change frequently, more frequent testing could help understand if students are moving in and out of food insecurity.
• Improved Income Measurement: Further research is needed to develop more accurate methods for measuring annual income among college students, considering their multiple and often inconsistent income streams.