WillTry Scale

Abstract

The WillTry Scale is an interviewer-administered psychometric instrument developed to evaluate children’s self-reported willingness to try diverse fruits and vegetables, encompassing both familiar (common) and unfamiliar (novel) varieties. Developed originally by Jessica L. Thomson and colleagues (2010) to support nutritional surveillance and intervention programming within health-disparate pediatric populations—notably in the rural Lower Mississippi Delta—the scale addresses a pivotal behavioral precursor to dietary variety and nutritional adequacy in youth aged 5 to 14 years. The standard operationalized battery consists of 22 target produce items categorized into fruit and vegetable domains, scored across a three-point Likert-type response scale: Wouldn't try (0 points), Would maybe try (1 point), and Would try (2 points). This yield generates a cumulative composite score ranging from 0 to 44, alongside distinct subscale scores reflecting domain-specific willingness (Fruit Willingness vs. Vegetable Willingness) and familiarity dimensions (Common Produce vs. Novel Produce). Psychometric evaluations demonstrate strong internal consistency reliability, with Cronbach’s alpha values typically ranging from α = .84 to .91 across composite and subscale iterations, as well as satisfactory test-retest reliability across multi-week intervals (intraclass correlation coefficients ranging from .65 to .78). Criterion-related and predictive validities have been rigorously confirmed using objective, weighed plate-waste metrics from direct snack-feeding observations, verifying that children expressing higher self-reported willingness consume a statistically significant higher percentage of offered produce. By bridging subjective food neophobia assessment and actual consumption behavior, the WillTry Scale serves as an empirically robust evaluative tool for pediatric obesity prevention initiatives, school nutrition trials, and community public health interventions.

Keywords

WillTry Scale, pediatric nutrition, food neophobia, fruit and vegetable consumption, dietary assessment, willingness to try, child feeding behavior, nutritional psychometrics, Lower Mississippi Delta, obesity prevention

Authors

The WillTry Scale was developed by an interdisciplinary team of behavioral scientists, clinical nutritionists, and biostatisticians affiliated with the United States Department of Agriculture (USDA) Agricultural Research Service (ARS) and regional academic medical centers:

• Jessica L. Thomson, PhD – Research Biostatistician, United States Department of Agriculture, Agricultural Research Service (USDA-ARS), Southern Regional Research Center, Diet, Genomics, and Immunology Laboratory / Lower Mississippi Delta Nutrition Intervention Research Initiative (NIRI).
• Beverly J. McCabe-Sellers, PhD, RD – Research Nutritionist, USDA-ARS Lower Mississippi Delta NIRI, Little Rock, Arkansas.
• Evelyn Strickland, MS, RD – Project Coordinator and Dietary Specialist, USDA-ARS Lower Mississippi Delta NIRI.
• Denise Lovera, MS – Research Associate, USDA-ARS Lower Mississippi Delta NIRI.
• Heather J. Nuss, PhD – Behavioral Scientist, Department of Public Health, Louisiana State University Health Sciences Center.
• Kathleen Yadrick, PhD, RD – Professor and Chair, Department of Nutrition and Food Systems, The University of Southern Mississippi, Hattiesburg, Mississippi.
• Suzanne Duke, MS – Statistician, USDA-ARS.
• Margaret L. Bogle, PhD, RD – Executive Director, USDA-ARS Lower Mississippi Delta Nutrition Intervention Research Initiative, Little Rock, Arkansas.

Purpose

Pediatric overweight and childhood obesity represent severe public health crises with far-reaching physiological, metabolic, and psychosocial sequelae. Epidemiological research routinely underscores that a dietary pattern characterized by adequate consumption of nutrient-dense fruits and vegetables confers substantial protective effects against chronic metabolic diseases, including type 2 diabetes mellitus, cardiovascular dysfunction, hypertension, and systemic low-grade inflammation. Guidelines published by the American Academy of Pediatrics (AAP) and the World Health Organization (WHO) underscore that establishing diverse produce consumption during middle childhood is vital for long-term health trajectories. However, achieving recommended produce intake remains difficult, particularly among socioeconomically disadvantaged and rural populations where food insecurity, dietary deserts, and limited exposure converge to restrict dietary diversity.

A primary psychological and behavioral bottleneck governing produce consumption in children is the reluctance or refusal to taste unfamiliar or sensorially challenging foods, a phenomenon intimately connected to food neophobia and pickiness. While general neophobia instruments exist, they frequently rely upon retrospective parental reports or evaluate abstract personality-like traits rather than concrete behavioral intentions. Such proxy methodologies are prone to informant bias, social desirability, and inaccurate maternal or paternal estimation of a child's willingness when separated from the domestic dining context. The WillTry Scale was purposefully engineered as a child-centered, interviewer-administered instrument that directly measures a child’s willingness to try specific, visualizable fruits and vegetables.

The primary clinical and research purposes of the WillTry Scale include:

• Direct Behavioral Assessment: Bypassing parent-proxy distortions by asking children directly about their willingness to try individual foods represented via standardized, high-resolution photographic flash cards.
• Programmatic Outcome Evaluation: Serving as a sensitive pre- and post-intervention outcome metric in school-based, summer camp, and community nutrition education programs designed to expand dietary repertoire and combat pediatric obesity.
• Granular Construct Differentiation: Distinguishing between willingness toward sweet, sensorially accessible targets (fruits) versus typically more bitter, non-sweet, texturally variable targets (vegetables), as well as differentiating ubiquitous produce (common) from culturally or geographically exotic targets (novel).
• Targeting Pediatric Feeding Difficulties: Providing pediatricians, feeding therapists, and clinical dietitians with an empirical baseline of food rejection patterns to tailor systemic desensitization, sensory-based feeding therapies, and positive reinforcement protocols.

Psychological Construct

The WillTry Scale operationalizes the core psychological construct of behavioral willingness to ingest within the domain of pediatric ingestive behavior. Behavioral willingness, rooted in the Prototype Willingness Model, differs subtly yet critically from behavioral intention. Whereas an intention entails a premeditated, deliberate plan to enact a target behavior, willingness reflects an individual's openness to engaging in a behavior when provided with an immediate risk-free or low-cost opportunity within an enabling environment. In young children whose dietary selections are predominantly constrained by adult provision, behavioral willingness constitutes the most ecologically valid cognitive antecedent of actual food tasting.

The construct measured by the WillTry Scale is multi-faceted, reflecting interactive cognitive, affective, sensory, and familiarity dynamics. It is organized into distinct subscale dimensions:

1. Fruit Willingness Subscale

Fruit willingness captures a child's willingness to try items characterized by natural sweetness, sucrose/fructose composition, and relatively high juiciness (e.g., apples, bananas, strawberries, watermelon, and kiwifruit). Innate human evolutionary biology favors sweet gustatory inputs because sweetness historically signaled non-toxic, calorie-dense sustenance. Consequently, willingness scores for fruits systematically exhibit higher baseline means and lower refusal variance than vegetable scores. Diminished fruit willingness typically signals profound, generalized sensory defensiveness or heightened severe food neophobia.

2. Vegetable Willingness Subscale

Vegetable willingness quantifies openness toward items that frequently feature bitter notes, fibrous or astringent mouthfeels, sulfurous aromas, and low caloric density (e.g., broccoli, peas, asparagus, zucchini, and eggplant). Evolutionarily, bitter and astringent flavors served as warning indicators of plant-derived alkaloids and phytotoxins. As children possess higher densities of fungiform taste buds, their sensitivity to bitter chemical compounds (such as 6-n-propylthiouracil, or PROP) is heightened. Vegetable willingness thus acts as an informative barometer of a child’s susceptibility to sensory rejection and contextual food pickiness.

3. Common vs. Novel Produce Dimensions

Within both fruit and vegetable categories, the scale differentiates between common foods (ubiquitously available items such as carrots, potatoes, apples, and bananas) and novel foods (produce that is culturally, geographically, or socioeconomically rare within the target demographic, such as dragonfruit, papaya, kiwifruit, and eggplant). The novel food dimension isolates acute food neophobia—the fear of the unfamiliar—while the common food dimension identifies entrenched selective or picky eating, wherein a child refuses even heavily advertised, culturally normative items that they have likely encountered repeatedly.

Theoretical Framework

The development and interpretation of the WillTry Scale rest upon an intersection of developmental psychology, behavioral economics, evolutionary anthropology, and nutritional epidemiology. Three foundational frameworks govern its architecture:

1. Omnivore’s Dilemma and Evolutionary Food Neophobia

Originating from the seminal theoretical work of Paul Rozin (1976) and later expanded by Patricia Pliner (1994), the omnivore's dilemma posits that omnivorous species must balance two opposing drives: neophilia (an exploratory interest in novel food sources necessary for meeting diverse micronutrient demands) and neophobia (a cautious reluctance or fear of unverified substances that may contain lethal toxins). In human development, food neophobia exhibits a distinct ontogenetic trajectory. It is minimal during infancy and early complementary feeding, peaks sharply between 2 and 6 years of age—coinciding with increased locomotor independence from parental supervision—and gradually diminishes across middle childhood as cognitive appraisal and experience mature.

The WillTry Scale operationalizes this developmental transition. In populations characterized by limited dietary diversity, such as rural, low-income areas, the neophobic response often remains elevated because repeated positive taste exposures—the primary experiential mechanism that down-regulates neophobic fear through learned safety—are structurally absent.

2. The Prototype Willingness Model (PWM)

Formulated by Frederick X. Gibbons and Meg Gerrard, the Prototype Willingness Model acknowledges that much of youth health behavior does not stem from rational, reasoned action pathways, but rather through reactive, opportunistic decision-making pathways. The scale captures this intuitive reaction: when presented with a concrete visual prompt (e.g., a photograph of sliced kiwi or fresh broccoli), the child reacts to the behavioral prospect of tasting it. The 3-point categorical ladder (Wouldn't try → Would maybe try → Would try) maps directly onto the spectrum of willingness, tracing the child's transition from outright avoidance, to ambivalent contemplation, to behavioral readiness.

3. Social Cognitive Theory and Familiarity Mechanics

Albert Bandura's Social Cognitive Theory emphasizes reciprocal determinism: cognitive beliefs, environmental exposures, and overt behavior continuously interact. Research by Leann Birch and colleagues demonstrates that children require between 8 to 15 repeated, non-coercive sensory exposures to a novel food before familiarity is established and affective preference shifts toward acceptance. In the WillTry paradigm, photographic prompts establish cognitive familiarity without imposing immediate postingestive pressure, accurately mapping the child's current stage of behavioral readiness along the exposure continuum.

Validity

The WillTry Scale has been subjected to empirical validity evaluations across several pediatric cohorts, most prominently within the Lower Mississippi Delta Nutrition Intervention Research Initiative (LMD NIRI) programs in Arkansas and Mississippi (Thomson et al., 2010).

Predictive and Criterion-Related Validity

A challenge in pediatric psychometrics is the reliance on proxy reports or self-report recall methods that suffer from severe social desirability and memory biases. The validation of WillTry addressed this by using direct, weighed plate-waste feeding sessions as the external criterion. In experimental validation studies:

• Children aged 5 to 14 completed the WillTry survey, supported by standardized photographic flash cards.
• Over subsequent multi-week observation periods (e.g., 6-week summer day camp snack feeding programs), children were presented with 18 distinct pre-weighed fresh fruit and vegetable snacks (13 fruits, 5 vegetables) with precise gram portions measured using calibrated analytical scales.
• Uneaten refuse and containers were collected, re-weighed, and calculated as the consumed percentage of offered produce.
• Statistical analysis confirmed that higher WillTry composite scores significantly predicted greater actual consumption percentages (Spearman rank correlations ranging from $r_s = .35$ to $r_s = .52$, $p < .001$). Children who endorsed the “Would try” (2 points) option for specific items displayed significantly higher consumption of those identical items during snack periods compared to those who endorsed “Would maybe try” or “Wouldn't try.”

Construct and Convergent Validity

Construct validity was evaluated by examining relationships between WillTry scores and established parent-reported and child-reported eating dimensions:

• Picky Eating Indicators: Children identifying as less picky (reporting that they “will eat anything offered”) scored significantly higher on both total WillTry ($M = 36.4, SD = 5.2$) and vegetable-specific subscales than children categorized as highly selective eaters ($M = 28.1, SD = 6.8$, $p < .01$).
• Self-Perceived Healthy Eater Status: Children who categorized themselves as “healthy eaters” exhibited higher willingness scores relative to those reporting “sometimes” or “no” ($F(2, 142) = 8.74, p < .001$).
• Convergence with Food Neophobia: Inverse correlations were established between the WillTry total score and standardized pediatric food neophobia inventories (such as the Child Food Neophobia Scale, CFNS), demonstrating convergent validity ($r = -.58, p < .001$).

Discriminant Validity

Discriminant validity was established by comparing child responses to non-food behavioral items and demographic controls (such as sibling count), which showed non-significant correlations with willingness metrics ($r < .08, p > .30$). Furthermore, the scale discriminated well between high-sugar common fruits (exhibiting ceiling effects in willingness) and novel bitter vegetables (exhibiting wide individual variance).

Reliability

The reliability profile of the WillTry Scale demonstrates robust psychometric stability and internal consistency across diverse school-aged demographics:

Internal Consistency Reliability

Across validation studies conducted by Thomson et al. (2010) involving pediatric samples in Arkansas (AR) and Mississippi (MS1, MS2), internal consistency was computed using Cronbach's coefficient alpha (α):

• Total Scale (22 items): Cronbach's α values ranged from .84 to .91, reflecting high homogeneity among items while retaining sufficient breadth across produce categories.
• Fruit Willingness Subscale: Demonstrated strong internal consistency, with α values between .79 and .86.
• Vegetable Willingness Subscale: Consistently demonstrated high internal consistency, with α values between .81 and .88, confirming that willingness toward diverse vegetables forms a cohesive behavioral dimension.
• Item-Total Correlations: Corrected item-total correlation coefficients remained consistently positive, with most items exceeding $r_{it} = .38$, confirming that no single item undermined scale integrity.

Test-Retest Reliability and Temporal Stability

Temporal stability was evaluated across a 6-week test-retest interval among summer camp cohorts who received no interim nutritional desensitization. The intraclass correlation coefficient (ICC) for the total WillTry composite score was ICC = .74 (95% CI [.63, .82]), demonstrating acceptable-to-good stability over an extended developmental timeframe. Subscale test-retest correlations yielded ICCs of .71 for fruits and .68 for vegetables, indicating that in the absence of targeted intervention, children's reported willingness to try produce remains remarkably stable across middle childhood.

Factor Analysis

The structural dimensionality of the WillTry Scale has been explored using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to determine whether willingness to try produce operates as a single global trait or a multidimensional structure.

Exploratory Factor Structure

Initial principal axis factoring with promax (oblique) rotation supported a clear two-factor conceptualization separating items along botanical and sensory lines:

• Factor 1: Vegetable Willingness. This dominant factor accounted for approximately 28.4% to 33.1% of the total variance. High-loading items (λ > .50) included broccoli, peas, asparagus, green beans, zucchini, and eggplant.
• Factor 2: Fruit Willingness. Accounting for an additional 14.2% to 18.6% of the variance, this factor comprised items characterized by sweetness and high moisture content, including strawberries, watermelon, grapes, apples, and bananas.

A secondary hierarchical analysis further distinguished items across a crossed familiarity axis, yielding a four-factor model: Common Fruits, Novel Fruits, Common Vegetables, and Novel Vegetables. The correlation between the latent Fruit and Vegetable factors remained moderate ($r = .42$ to $.51$), supporting the use of both distinct subscale scores and a broad composite score.

Confirmatory Factor Analysis and Model Fit

CFA conducted on subsequent independent samples verified the superiority of the two-factor correlated model (Fruit vs. Vegetable) over a single-factor unidimensional model. Standard structural equation modeling indices confirmed good model fit:

• Comparative Fit Index (CFI): .934 to .951 (exceeding the standard ≥ .90 benchmark).
• Tucker-Lewis Index (TLI): .926 to .944.
• Root Mean Square Error of Approximation (RMSEA): .048 (90% CI [.038, .058]), well below the acceptable .06 threshold.
• Standardized Root Mean Square Residual (SRMR): .052.

Standardized factor loadings across items were robust, ranging from λ = .44 to .78 for vegetable items and λ = .41 to .72 for fruit items, confirming that all 22 target foods contribute reliably to their respective latent constructs.

Instrument / Measurement Tool

• Instrument Name: WillTry Scale
• Target Population: Children and early adolescents aged 5 to 14 years
• Administration Format: Individual, one-on-one interviewer-administered survey assisted by standardized visual aids (photographic flash cards depicting raw, whole, cut, or sliced produce)
• Item Count: 22 standardized target food items
• Subscale Structure:
  • Fruit Willingness Subscale: 11 items (Apples, Bananas, Oranges, Strawberries, Grapes, Kiwifruit, Watermelon, Pears, Mango, Papaya, Dragonfruit)
  • Vegetable Willingness Subscale: 11 items (Carrots, Peas, Broccoli, Corn, Potatoes, Green beans, Tomatoes, Capsicum/Bell peppers, Asparagus, Eggplant, Zucchini)
  • Cross-cutting Dimension: Common Produce vs. Novel Produce
• Response Scale: 3-point categorical response scale:
  • Wouldn't try = 0 points
  • Would maybe try = 1 point
  • Would try = 2 points
• Scoring Rules:
  • All items are positively scored; there are no reverse-scored items.
  • Total WillTry Composite Score: Sum of all 22 items, ranging from 0 to 44.
  • Fruit Willingness Subscale Score: Sum of items 1–8 and 17–19, ranging from 0 to 22.
  • Vegetable Willingness Subscale Score: Sum of items 9–16 and 20–22, ranging from 0 to 22.
  • Score Interpretation: Higher numerical scores denote greater openness and behavioral willingness to try fruits and vegetables. Low scores (< 22 total) indicate pronounced food neophobia or heightened feeding selectivity.
• Administration Duration: Approximately 8 to 12 minutes per child.

Permissions & Fee and Test Year

The WillTry Scale was developed in 2006–2008 and formally published in peer-reviewed literature in 2010 by Jessica L. Thomson and co-authors under the auspices of the United States Department of Agriculture (USDA) Agricultural Research Service (ARS). Because it was developed through federally funded research conducted by USDA-ARS scientists, the core assessment instrument is considered to be in the public domain and may be utilized for non-commercial academic, educational, and clinical research purposes without licensing fees.

Investigators wishing to adapt the tool are advised to maintain standardized photographic stimulus protocols and cite the primary validation paper (Thomson et al., 2010). Inquiries regarding original high-resolution stimulus materials, administration manuals, or regional adaptations may be directed to the corresponding author, Dr. Jessica L. Thomson, at the USDA-ARS.

References

• Birch, L. L., & Marlin, D. W. (1982). I don't like it; I never tried it: Effects of exposure on two-year-old children's food preferences. Appetite, 3(4), 353–360. https://doi.org/10.1016/S0195-6663(82)80053-6
• Cullen, K. W., Watson, K., & Zakeri, I. (2008). Improvements in middle school students' diet with the Texas Public School Nutrition Policy. American Journal of Public Health, 98(1), 111–117. https://doi.org/10.2105/AJPH.2007.113944
• Gibbons, F. X., Gerrard, M., Blanton, H., & Russell, D. W. (1998). Reasoned action versus social reaction: Thinking about health behaviors in adolescents and young adults. In The Reasoned Action Approach to Health Risk Behavior (pp. 45–75). Lawrence Erlbaum Associates.
• Guthrie, C. A., Rapoport, L., & Wardle, J. (2000). Young children's food preferences: A comparison of three types of food. Health Education & Behavior, 27(6), 759–777. https://doi.org/10.1177/109019810002700609
• Loewen, R., & Pliner, P. (2000). Effects of prior exposure to contextual cues on children's willingness to try unfamiliar foods. Appetite, 35(1), 81–88. https://doi.org/10.1006/appe.2000.0336
• Pliner, P. (1994). Development of measures of food neophobia in children. Appetite, 23(2), 147–163. https://doi.org/10.1006/appe.1994.1043
• Rozin, P. (1976). The selection of foods by rats, humans, and other animals. In J. S. Rosenblatt, R. A. Hinde, E. Shaw, & C. Beer (Eds.), Advances in the Study of Behavior (Vol. 6, pp. 21–76). Academic Press. https://doi.org/10.1016/S0065-3454(08)60081-9
• Rubio, B., Rigal, N., Boireau-Ducept, N., Mallet, P., & Meyer, T. (2008). Measuring food neophobia in children: Validation of a French version of the Food Neophobia Scale. Appetite, 50(2–3), 566. https://doi.org/10.1016/j.appet.2007.10.010
• Thomson, J. L., McCabe-Sellers, B. J., Strickland, E., Lovera, D., Nuss, H. J., Yadrick, K., Duke, S., & Bogle, M. L. (2010). Development and evaluation of WillTry: An instrument for measuring children's willingness to try fruits and vegetables. Appetite, 54(3), 465–472. https://doi.org/10.1016/j.appet.2009.12.015
• Tussing-Humphreys, L., Thomson, J. L., McCabe-Sellers, B., Strickland, E., Lovera, D., & Bogle, M. (2012). A school-based fruit and vegetable snacking pilot intervention for Lower Mississippi Delta children. Infant, Child, & Adolescent Nutrition, 4(5), 295–302. https://doi.org/10.1177/1941406412455986

13. Items of the Scale (Questionnaire)