henhao
Sign in
BioengineeringFood Research International
Peer-reviewed journal article

The Meatball Matchup: Plant vs. Animal Proteins on Campus

Skyler R. St. PierreiD, Aeneas O. KoosisiD, Nancy Ruonan ZhangiD, Ellen Kuhl AiD

Abstract

Institutional dining halls increasingly serve plant-based meat products, yet it remains unclear how consumers perceive these products relative to animal-based options in real-world settings. This study compares consumer sensory perceptions of two plant-based meatballs (soy, soy-wheat) and two animal-based meatballs (beef, beef-mushroom) among university dining hall patrons (n = 116), complemented by instrumental Texture Profile Analysis. Animal-based meatballs received significantly higher ratings for moistness, meatiness, fattiness, and tastiness (all p < 0.001), with the meatiness gap being the largest ({Delta} = 1.40 on a 5-point scale). Texture analysis revealed that animal-based samples were significantly harder, more cohesive, and chewier than plant-based samples. In contrast, consumers perceived no differences in chewiness or hardness between categories, revealing a disconnect between instrumental and sensory measures. Just-About-Right penalty analysis identified insufficient savoriness as a universal improvement target across all products, including beef. Flavor and texture emerged as the dominant drivers of dining choice, while sustainability and animal welfare ranked lowest in importance. These findings suggest that sensory parity--particularly in moistness, meatiness, and savoriness--may drive acceptance of plant-based meat more strongly than sustainability messaging. Data and code are available at https://github.com/LivingMatterLab/AI4Food

PDFDOICodeData
Full-text digest ↓

02 /Specification & Classification

Click any tag to browse matching papers
Citations
5
FWCI
15.32
Field percentile
top 1%
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Opus 5.5 · xhigh effort · 7 Nov 2026 · from the full text of version 2 (10 May 2026) · CC BY 4.0

Stanford diners (n = 116) rated soy and soy-wheat meatballs below beef-based ones on moistness, meatiness, fattiness and tastiness, yet perceived no hardness or chewiness gap despite instrumental differences.

Most sensory studies of plant-based meat take place in controlled laboratories, and it is not known how far instrumental texture differences carry over into what consumers perceive. This work tests both questions in a working university dining hall, pairing consumer ratings, Just-About-Right (JAR) penalty analysis and choice-factor importance with Texture Profile Analysis (TPA) of the same four meatballs.

Key results

  1. ShownIn this sample of diners, both beef-based meatballs scored higher than both soy-based meatballs on moistness, meatiness, fattiness and tastiness. Meatiness showed the widest gap.

    All p < 0.001. Meatiness Δ = 1.40 and moistness Δ = 0.85 on a 5-point scale. Share of participants (n = 116) scoring the animal meatballs higher: 71% for moistness, 83% for meatiness, 84% for fattiness and 77% for tastiness, with 11% or fewer rating no difference. Friedman tests found differences across the four products on all eight attributes (W = 0.12 to 0.38). · Abstract; 3.2. Consumer Sensory Attribute Ratings; Animal meatballs outperform plant meatballs in…

  2. ShownInstrumental texture split cleanly by category. On all six TPA parameters, both beef meatballs differed significantly from both plant meatballs, and neither category showed differences within itself.

    Stiffness: Beef-Mushroom 101.7 kPa, Beef 91.7 kPa, Soy-Wheat 56.2 kPa, Soy 45.4 kPa. Chewiness in the same order: 0.57 N, 0.56 N, 0.19 N, 0.09 N. One-way ANOVA with Tukey HSD, p < 0.05. Samples were compressed twice to 50% strain at 25%/s; n = 37 cylindrical cores were analysed after 3 of 40 were excluded for slippage. · 2.5. Texture Profile Analysis; 3.4. Texture profile analysis

  3. SuggestedInstrumental and perceived texture did not match. Diners saw no category gap in hardness or chewiness, and TPA rankings did not predict the matching sensory ratings. TPA values instead ranked products in line with perceived meatiness, though not significantly.

    Kendall τ = 0.00 (p = 1.00) for TPA vs sensory hardness and for TPA vs sensory chewiness. Instrumental hardness: 2.3-2.6 N for beef vs 1.1-1.4 N for plant samples. Soy-Wheat: 1.41 N instrumental hardness yet highest perceived hardness (3.56/5). Chewiness: 41% rated animal higher, 41% plant higher, 18% no difference. TPA vs meatiness τ = +0.33 to +0.67; none significant with n = 4 products. · 3.2. Consumer Sensory Attribute Ratings; 3.5. Texture-Sensory Correlation Analysis; Mechanical features fail to predict sensory…

  4. ShownJAR penalty analysis found that too little savoriness lowered tastiness for every product, beef included. The plain Beef meatball carried the largest savoriness penalty.

    Penalties for savoriness rated 'not enough' ranged from 0.71 (Soy-Wheat) to 1.36 (Beef) on the 5-point tastiness scale; the actionable threshold was ≥0.50. Soy-Wheat had the largest JAR deviations (too little fattiness, savoriness and moistness; too much chewiness) and was penalized for chewiness in both directions. 31% rated Soy's fibrousness 'not enough'. · 2.6. Statistical Analysis; 3.3. Just-About-Right Penalty Analysis; None of the meatballs were savory enough

  5. ShownWhen choosing between animal and plant options, diners rated flavor and texture far above sustainability and animal welfare. Neither these priorities nor meat attachment tracked tastiness or sensory ratings.

    Mean importance on a 0-100 slider: flavor 82, texture 65, familiarity 60, health 53, availability 50, climate sustainability 28, animal welfare 25, social norms 24. Friedman χ2(7) = 313.07, p < 0.001, W = 0.349. Meat attachment vs tastiness: all p > 0.05. Choice factors vs sensory ratings: |ρ| < 0.25. The only gender effect: males rated Soy-Wheat higher (U = 2496.0, p = 0.015). · 3.6. Choice Factor Importance; 3.7. Demographic and Individual Difference…

Methods, data and reproducibility

Field survey during a single lunch service at a Stanford dining hall. 116 diners rated two plant-based meatballs (Soy, Soy-Wheat) and two beef-based meatballs (Beef, Beef-Mushroom) on eight 5-point attributes and five 3-point JAR scales, plus meat attachment and choice-factor sliders. Cored samples were tested by double-compression TPA on a rheometer, and the product rankings were correlated with sensory scores.

  • Dining hall patrons who completed the sensory survey for all four meatballs (Stanford University, single lunch service) · human · n = 116
  • Four meatball products: Soy and Soy-Wheat (commercial plant-based brands), Beef and Beef-Mushroom (prepared in-house by Stanford Dining) · n = 4
  • Cylindrical meatball core samples (8 mm diameter, about 10 mm high) for Texture Profile Analysis · n = 40 prepared (10 per product); 37 analysed
  • Open-ended participant comments coded by thematic analysis · human · n = 98
Methods & tools
Qualtrics, Discovery HR 20 Rheometer (TA Instruments), Texture Profile Analysis (TPA), Just-About-Right (JAR) scales with penalty analysis, Meat Attachment Questionnaire (MAQ), 5-item dependence subset, custom Python scripts, Friedman test with Kendall's W, Wilcoxon signed-rank tests with Bonferroni correction, Mann-Whitney U and Kruskal-Wallis H tests, one-way ANOVA with Tukey HSD, Kendall's tau rank correlation
Code
https://github.com/LivingMatterLab/AI4Food
Data & accessions
https://github.com/LivingMatterLab/AI4Food
Reproducibility gaps
Python package versions and environment files are not given; the analysis is described only as custom Python scripts.
Limitations (6)
  • The sample is homogeneous (96% aged 18-24, mostly Stanford undergraduates), so results may not generalize to wider food-service populations. · stated by the authors
  • The real dining setting left variables uncontrolled (accompaniments, hunger, social context, self-selection), and the colored flags kept brands anonymous but revealed protein category, which may have biased expectations. · stated by the authors
  • TPA was run at room temperature rather than serving temperature. The single cross-sectional session cannot show whether repeated exposure would narrow the sensory gaps. · stated by the authors
  • The texture-sensory correlations rest on only 4 products, where Kendall's τ reaches p < 0.05 only with perfect agreement. The authors call these patterns suggestive and needing replication. · stated by the authors
  • The text does not say whether the order in which the four meatballs were tasted and rated was randomized or counterbalanced, so order effects cannot be excluded. · noted in this summary
  • Some values differ between sections: omnivores are 94% in Results vs 89.8% in Discussion; flavor importance is 82 vs 83/100; the TPA parameters listed at τ = +0.67 with meatiness differ; and stiffness is in N/mm in Methods but kPa in Results. · noted in this summary

Who should read:Read this if you study consumer acceptance of plant-based meat, or want a worked example that pairs field sensory ratings and JAR penalty analysis with rheometer-based Texture Profile Analysis of the same products.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISPierre '24Koosis '26Pierre '25Koosis '26Dunne '25Koosis '26
77% semantic matchBioengineering
Skyler R. St. Pierre, Ethan C. Darwin, Divya Adil+9 more·npj Science of Food·Nov 15, 2024· 43 citations