Limits...
Validation of the food access survey tool to assess household food insecurity in rural Bangladesh.

Na M, Gross AL, West KP - BMC Public Health (2015)

Bottom Line: All nine items were flagged with statistically significant DIF between key demographic-and socioeconomic subgroups (p < 0.001); however, none of the detected DIF was considered practically significant given small effect sizes (variance explained by group membership and interaction term < 1%).The total summed score over the polytomous FAST was inversely associated with household wealth, dietary diversity score and maternal body mass index, demonstrating external construct validity.Validation of this type of studies are recommended for similar Likert food insecurity scales.

View Article: PubMed Central - PubMed

Affiliation: Department of International Health, Johns Hopkins Bloomberg School of Public Health, Center for Human Nutrition, 615 N. Wolfe St., W2041, 21205, Baltimore, MD, USA. mna1@jhu.edu.

ABSTRACT

Background: Perception-based Likert scale are commonly used to assess household food insecurity. The aim of this study was to evaluate the psychometric properties and external construct validity of the 9-item Food Access Survey Tool (FAST) in a population-based randomized controlled trial.

Methods: Participating women (n = 11,992) were asked to recall the frequencies of nine food insecurity experiences over the past 6 months on a 5-point Likert scale. The Rasch partial credit model was used to study the item category severity and differential item functioning (DIF) by literacy status, respondents' age, land ownership and household sizes. Principal component analysis (PCA), non-parametric methods, and cumulative ordinal logistic regression models were applied to examine the Rasch model assumptions, namely unidimensionality, monotonicity and measurement invariance (non-DIF).

Results: All items demonstrated good model fit with acceptable values of fit statistics (infit). PCA as well as other indices (Cronbach's alpha = 0.85, scalability coefficient = 0.48) indicated that all items fit in a single statistical dimension. The ordered responses of nine items displayed monotonic increasing item category severity as expected theoretically. All nine items were flagged with statistically significant DIF between key demographic-and socioeconomic subgroups (p < 0.001); however, none of the detected DIF was considered practically significant given small effect sizes (variance explained by group membership and interaction term < 1%). The total summed score over the polytomous FAST was inversely associated with household wealth, dietary diversity score and maternal body mass index, demonstrating external construct validity.

Conclusion: The polytomous FAST is internally and externally valid tool to measure household food insecurity in rural Bangladesh. Validation of this type of studies are recommended for similar Likert food insecurity scales.

No MeSH data available.


Related in: MedlinePlus

Distribution of estimated household food insecurity latent score and item category severity (Thurstonian thresholds): R, Rarely; S, Sometimes; O, Often; M, Mostly
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4561472&req=5

Fig2: Distribution of estimated household food insecurity latent score and item category severity (Thurstonian thresholds): R, Rarely; S, Sometimes; O, Often; M, Mostly

Mentions: In Fig. 2, the histogram on the top represents the distribution of households by their food insecurity latent score on a logit scale. The high bar to the far left represented the 46.7 % households that had reported no food insecurity related experiences to the scale. On the bottom of Fig. 2, item category severity (Thurstonian threshold) of FAST items were reordered by the category severity of the “rarely” response. Item severity estimates for the “rarely”, “sometime”, “often”, and “mostly” category were derived from the step-wise PCM and displayed increasing severity from “rarely” to “mostly” for all items. Estimates of category severity were also dispersed along the logit scale of food insecurity latent trait as expected. Logit is the log-odds transformation of the probability of affirming a given category of a scale item. The logit value is expected to be low when the item response category reflects mild food insecurity and high when the category reflects more severe food insecurity. In Fig. 2, the category severity in logit for “rarely” ranged from 0.48 in item 7 “purchase rice often” to 2.49 in item 2 “have to eat other grains”. This means households employed different coping strategies at different level of insecurity: when food insecurity was relatively mild, households reallocate resources to purchase rice more often with smaller amount each time; when food insecurity became worse, household members started to sacrifice their amount of food consumed by eating less (item 4), having no square meals (item 1 reversed), skipping the entire meal (item 3) and even changing their usual grain preference (item 2). Social acceptable strategies in food acquisition (item 9 “borrow food” and item 8 “take food on credit”) were taking place in between the coping strategies of consumption modification. Contrary to the theoretical expectation, however, worrying about food (item 6) did not happen at early stage of food insecurity reflected by a relative higher item severity.Fig. 2


Validation of the food access survey tool to assess household food insecurity in rural Bangladesh.

Na M, Gross AL, West KP - BMC Public Health (2015)

Distribution of estimated household food insecurity latent score and item category severity (Thurstonian thresholds): R, Rarely; S, Sometimes; O, Often; M, Mostly
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4561472&req=5

Fig2: Distribution of estimated household food insecurity latent score and item category severity (Thurstonian thresholds): R, Rarely; S, Sometimes; O, Often; M, Mostly
Mentions: In Fig. 2, the histogram on the top represents the distribution of households by their food insecurity latent score on a logit scale. The high bar to the far left represented the 46.7 % households that had reported no food insecurity related experiences to the scale. On the bottom of Fig. 2, item category severity (Thurstonian threshold) of FAST items were reordered by the category severity of the “rarely” response. Item severity estimates for the “rarely”, “sometime”, “often”, and “mostly” category were derived from the step-wise PCM and displayed increasing severity from “rarely” to “mostly” for all items. Estimates of category severity were also dispersed along the logit scale of food insecurity latent trait as expected. Logit is the log-odds transformation of the probability of affirming a given category of a scale item. The logit value is expected to be low when the item response category reflects mild food insecurity and high when the category reflects more severe food insecurity. In Fig. 2, the category severity in logit for “rarely” ranged from 0.48 in item 7 “purchase rice often” to 2.49 in item 2 “have to eat other grains”. This means households employed different coping strategies at different level of insecurity: when food insecurity was relatively mild, households reallocate resources to purchase rice more often with smaller amount each time; when food insecurity became worse, household members started to sacrifice their amount of food consumed by eating less (item 4), having no square meals (item 1 reversed), skipping the entire meal (item 3) and even changing their usual grain preference (item 2). Social acceptable strategies in food acquisition (item 9 “borrow food” and item 8 “take food on credit”) were taking place in between the coping strategies of consumption modification. Contrary to the theoretical expectation, however, worrying about food (item 6) did not happen at early stage of food insecurity reflected by a relative higher item severity.Fig. 2

Bottom Line: All nine items were flagged with statistically significant DIF between key demographic-and socioeconomic subgroups (p < 0.001); however, none of the detected DIF was considered practically significant given small effect sizes (variance explained by group membership and interaction term < 1%).The total summed score over the polytomous FAST was inversely associated with household wealth, dietary diversity score and maternal body mass index, demonstrating external construct validity.Validation of this type of studies are recommended for similar Likert food insecurity scales.

View Article: PubMed Central - PubMed

Affiliation: Department of International Health, Johns Hopkins Bloomberg School of Public Health, Center for Human Nutrition, 615 N. Wolfe St., W2041, 21205, Baltimore, MD, USA. mna1@jhu.edu.

ABSTRACT

Background: Perception-based Likert scale are commonly used to assess household food insecurity. The aim of this study was to evaluate the psychometric properties and external construct validity of the 9-item Food Access Survey Tool (FAST) in a population-based randomized controlled trial.

Methods: Participating women (n = 11,992) were asked to recall the frequencies of nine food insecurity experiences over the past 6 months on a 5-point Likert scale. The Rasch partial credit model was used to study the item category severity and differential item functioning (DIF) by literacy status, respondents' age, land ownership and household sizes. Principal component analysis (PCA), non-parametric methods, and cumulative ordinal logistic regression models were applied to examine the Rasch model assumptions, namely unidimensionality, monotonicity and measurement invariance (non-DIF).

Results: All items demonstrated good model fit with acceptable values of fit statistics (infit). PCA as well as other indices (Cronbach's alpha = 0.85, scalability coefficient = 0.48) indicated that all items fit in a single statistical dimension. The ordered responses of nine items displayed monotonic increasing item category severity as expected theoretically. All nine items were flagged with statistically significant DIF between key demographic-and socioeconomic subgroups (p < 0.001); however, none of the detected DIF was considered practically significant given small effect sizes (variance explained by group membership and interaction term < 1%). The total summed score over the polytomous FAST was inversely associated with household wealth, dietary diversity score and maternal body mass index, demonstrating external construct validity.

Conclusion: The polytomous FAST is internally and externally valid tool to measure household food insecurity in rural Bangladesh. Validation of this type of studies are recommended for similar Likert food insecurity scales.

No MeSH data available.


Related in: MedlinePlus