Increasing Consumer Acceptance of Baby Leafy Greens Grown in a Controlled Environment

Progress report for GNE19-212

Project Type: Graduate Student
Funds awarded in 2019: $15,000.00
Projected End Date: 07/31/2022
Grant Recipient: Rutgers, The State University of New Jersey
Region: Northeast
State: New Jersey
Graduate Student:
Faculty Advisor:
Dr. Beverly Tepper
Rutgers University
Expand All

Project Information

Project Objectives:

The overall goal of this project is to develop a comprehensive approach to model consumer acceptance and purchase intent of different varieties of baby leafy greens. This model will allow us to identify consumer segments that differ in their expectations and preferences for baby leafy greens. This knowledge will help CEA operators better understand their consumers and assist them in selecting what varieties to grow. The strategies developed in this research can be translated and used in a variety of controlled agriculture settings. This project consists of two parts: a pilot study (Phase I) and a large consumer study (Phase II). Phase I Researchers have attempted to understand the links between consumer acceptance of leafy greens and personal characteristics with limited success [13], potentially due to the complexity of consumer behavior. Previous studies have generally examined only one or two domains in isolation to predict this multi-dimensional behavior.

The objectives of Phase 1 are:

1.To pilot test the use of questionnaires and methods that are most strongly related to consumer acceptance of different varieties of baby leafy greens across the three domains of interest; taste, demographics and personality/attitudes.

2.To identify and select the best questionnaires and sub-scales for use in Phase II. This step is critical to the success of Phase II as the administration of a large battery of questionnaires is not feasible in large consumer studies. Phase II It is well-known that individual consumers have different food preferences [4, 15]; these differences create a need to identify sub-groups of consumers who prefer different varieties of leafy greens. Under Phase II, we will conduct a large-scale consumer study in a community setting.

The objectives for Phase II are:

1.To utilize selected questionnaires and methods identified in Phase I to assess consumer liking of different varieties of baby leafy greens.

2.To apply multi-variate statistics to model consumer liking and purchase intent of baby leafy greens and identify consumer segments.


The purpose of this project is to understand factors affecting consumer acceptance and preference for baby leafy greens grown via vertical farming in a controlled indoor environment to guide selection of varieties by indoor farmers.

Interest in controlled environment agriculture (CEA) is growing rapidly due to concerns about feeding a growing global population. CEA methods such as aeroponics are highly sustainable, using up to 95% less water than traditional farming, and are ideal for growing in the urban or peri-urban settings that can be found throughout the Northeast region of the United States.

The current high demand for value-added crops that are sustainable, pesticide-free, and locally-sourced has attracted the attention of both entrepreneurs and investors. As a relatively young technology, there are many unanswered questions about vertical farming, such as processing conditions, efficient supply chain management, and consumer acceptance of the final product.

The question of consumer demand for leafy greens is not easily answered. The primary reason why a consumer will accept or reject a food is because of its taste. Bitter or sulfurous tastes in leafy greens and other vegetables can be off-putting to many people. However, many of these bitter compounds have nutritional or other health-promoting benefits, which is also of interest to the consumer. Individual differences in perception and preference for bitterness increase the need for variety in the marketplace. The ability to predict the type of greens desired by specific consumer groups is especially beneficial for CEA farmers, since they can consistently and reliably alter the environment in which their products are grown.

Peer-reviewed and publicly available research on the market potential and consumer acceptance of leafy greens produced through vertical farming is not available. Studies on consumption habits for bitter vegetables focus on demographic factors, whereas others emphasize taste or attitudinal factors. No studies have simultaneously examined all three of these domains in the same investigation to obtain a more comprehensive and holistic view of consumer acceptance. 

To accomplish this research, we are partnering with AeroFarms, a leader in aeroponic farming technology, whose world headquarters is located in urban Newark, NJ. We are currently conducting a collaborative project with AeroFarms funded by a grant from the Foundation for Food and Agricultural Research (FFAR). Preliminary data from that project has informed the consumer acceptance studies planned in this project. The knowledge generated in this project will inform and guide the planting programs and variety selections of a range of CEA operators.


Click linked name(s) to expand/collapse or show everyone's info
  • Dr. Edward Harwood (Researcher)
  • Dr. Jon Van Wagenen (Researcher)


Materials and methods:

Left: Baby leafy greens being prepared for participants. Middle: Graduate Student Regina O'Brien prepares baby leafy greens for testing. Right: Graduate Student Fay Alwattari serves samples to participants.

All research activities were approved by Rutgers University Institutional Review Board (Protocol#2019001829).

 I. Consumer Taste Research

  1. Participants:

1a. Participant Recruitment: Potential participants for consumer testing were recruited from the Rutgers University-New Brunswick campus through flyers and university email distribution lists. Interested participants were screened for allergies and consumption frequency of leafy greens before being selected for the study. Potential participants who did not consume leafy greens at least a few times per year were excluded from the study.


1b. Participant Demographics: A total of seventy-four consumers participated in the study. All participants consumed leafy green vegetables at least once per month and had no allergies or dietary restrictions.


  1. Testing Environment: All testing was done under white lighting and standard conditions in the Sensory Evaluation Laboratory at Rutgers University (New Brunswick, NJ). Participants completed both sessions in individual testing booths. Data was collected on computers equipped with RedJade software (Redwood City, CA).


  1. Selection of Plant Varieties:


3a. Selection of Flavor Profiles: Prior to the current experiment, we carried out a ‘Napping’ experiment with consumers to understand consumers’ general perception of baby leafy greens. This Napping experiment determined three areas in which consumers had grouped varieties of greens: 1. “Spicy/Pungent,” 2. “Bitter/Astringent,” and 3. “Sweet/Mild” types. These groups served as the basis for varietal selection and grouping.


3b. Varietal Selection: To reduce the amount of fatigue from tasting, a total of six varieties of baby leafy greens were selected, with two representative varieties from each category mentioned above (i.e. “spicy,” “bitter,” and “sweet”). Selection was based on product availability and previous use in the Napping experiment. Varieties were then classified as “spicy,” “bitter,” or “sweet” based on the results of their grouping in the Napping experiment (where applicable) and an informal discussion about the appropriate classification for each.


  1. Plant Material: All baby leafy green varieties were grown by AeroFarms (Newark, NJ) in an indoor aeroponic growing environment. Plants were harvested and delivered to Rutgers University one day prior to testing. Plants were stored in a refrigerator at 4C until being used for testing.

The following varieties are listed in Table 1 with their flavor profile classification.


Variety (Common Name)

Flavor Profile






AeroFarms (Newark, NJ)

Ruby Streaks


Green Kale


Red Kale


Red Romaine


Pac Choi



Baby leafy greens were stored in bowls and mixed with vinaigrette dressing (1/5c. white wine vinegar in 1c. olive oil) before being served. Greens were replaced every hour to avoid wilting. Greens were served in filled 12oz bowls labeled with random three-digit codes to prevent bias. Greens were served one-at-a-time in a balanced randomized order and all participants tasted all 6 varieties. Participants waited 1 minute between samples and were given filtered water (Poland Spring) and unsalted crackers (Nabisco) to cleanse their palate between samples.


  1. Questionnaires:

5a. Liking Questionnaires: Panelists evaluated all 6 samples on a 7-point hedonic (liking) questionnaire form ranging from ‘Do not like at all’ to ‘Like very much.’  Participants also rated their likelihood to purchase each sample on a 7-point scale ranging from ‘Not at all likely’ to ‘Extremely likely.’ Questions included liking of appearance, aroma, flavor, texture, and overall liking of each sample. At the end of each sample questionnaire, an optional write-in ‘Comments’ section was provided.


5b. Personality Questionnaires: Participants completed a series of questionnaires on their behaviors and personality traits. All questionnaires selected have been shown to provide insight into consumer food choice and preference. Participants completed the Three Factor Eating Questionnaire, the Food Adventurousness Scale, the Variety Seeking Questionnaire, and a demographic questionnaire.


  1. PROP Testing: Participants were tested for their sensitivity to bitter taste using the filter paper method developed by Zhao, Kirkmeyer, and Tepper(2003). Participants tasted a piece of filter paper impregnated with 6-n-propylthiouracil (PROP) and rated the intensity of the perceived taste on the Labeled Magnitude Scale. Participants were trained on the use of the LMS with a series of weights prior to evaluating NaCl and PROP filter paper discs. The ability to taste PROP has been studied for its influence on consumption and liking of bitter foods such as cruciferous vegetables and has been implicated in the role of preference for spicy and pungent foods.



  1. Data Collection

Participants completed two 30-minute sessions in the Sensory Evaluation Laboratory at Rutgers University. Participants were consented at the beginning of Session 1 before the study procedures. Participants were compensated with a $15 VISA gift card after the completion of both sessions.

1a. Session 1

Participants were consented at the beginning of Session 1. In this session, participants were presented with each of the 6 samples of baby leafy greens, one at a time in a randomized order. Participants tasted each sample and rated the liking of appearance, aroma, flavor, texture, overall liking, and purchase likelihood on the provided scales. Participants waited 1 minute between samples and were instructed to cleanse their palate between samples.

1b. Session 2

In Session 2, participants completed the TFEQ, VARSEEK, Food Adventurousness Scale, and demographic questionnaire. After the completion of the questionnaires, participants completed training on the use of the Labeled Magnitude Scale. Training was done by handing participants a series of weights and estimating the weight of each bottle on the LMS. Following training, participants were tested for their ability to taste PROP using the filter paper method described above. Participants waited one minute between all samples and were instructed to cleanse their palate with filtered water between NaCl and PROP filter paper discs.

  1. Data Analysis

Data analysis is currently being completed for Phase I data. Data will be analyzed in SAS (Cary, NC) and XLSTAT (Addinsoft; Paris, France). Mean differences in liking ratings across all samples will be done using Analysis of Variance. Hierarchical Cluster Analysis will be performed to cluster consumers based on liking ratings for each variety. The results from Phase I will be used to decide the appropriate questionnaires to be used in a larger (n=300) consumer study on a similar sample set.


II. Research in 2020-2022: Online Consumer Testing

Due to the COVID-19 pandemic, research at Rutgers University was halted for much of 2020. We are currently analyzing our data from the first study for publication. An online consumer test was developed which will has similar outcomes to those described in the consumer test, however due to the nature of online testing, there is a larger focus placed on visual sensory attributes and determinants for purchasing.  Data for this section of the online consumer study is currently being collected and is described in the following sections.


1. Participants

1a. Participant Recruitment: Four hundred regular consumers of leafy greens (consume 1x per month or more) will be recruited from a national sample. All participants will be aged 18 or older.


2. Online Survey

2a. Conjoint Analysis: A conjoint analysis approach was used to determine which sensory attribute are most likely to influence purchase: color (green or combination green/purple), texture (smooth texture or combination smooth/curly texture), and flavor quality (sweet or spicy). Visual attributes were selected based on work from Swegarden et al, 2019 on kale hybrids, in which some consumer segments were found to have high liking scores for novel colors (i.e. purple/red) and textures (i.e. curly). Flavor descriptors were selected as containing large differences in perception based on published work from our lab (O'Brien, Simon & Tepper, 2021).

In the conjoint survey, each combination of the three sensory attributes (color, texture, and flavor) are presented in a forced-choice question. Two combinations are presented at a time, and consumers must choose which product they are more likely to purchase.

2b. Bitter Food Liking: Consumers will be asked about their liking of a list of a variety of bitter and non-bitter foods. This data will be used as a proxy for PROP testing to understand more about consumers' liking of bitter vegetables. Consumers will be asked to rate their liking on a 9-point scale ranging from 'Dislike Extremely' to 'Like Extremely.'

2c. Maximum Differential (MaxDiff) Scaling: Maximum differential scaling will be used to determine which front-of-package marketing terms are most likely and least likely to influence their purchase of leafy greens. Terms include those related to nutritional content (i.e. a good source of 'Vitamin C', 'Antioxidants'), and those related to cultivation (i.e. 'certified organic', 'pesticide-free', 'farmed indoors').

2d. Demographic Information: Consumers will be asked which leafy greens they consume most frequently, and which sensory characteristics of leafy greens they like and dislike. Participants will be asked demographic information about their purchasing habits, age, income level, and ethnicity.


3. Data Analysis

3a. Conjoint and MaxDiff questionnaire analysis: For the conjoint analysis, individual utility scores will be calculated by Hierarchical Bayesian estimation, and analysis of variance (ANOVA) will be used to determine significant differences between levels and attributes. Cluster analysis of utility scores will be analyzed using Hierarchical Cluster Analysis. MaxDiff scores will also be evaluated with Hierarchical Bayesian estimation. Data from the demographic and bitter taste liking questions will be incorporated to model the reasons for consumer liking and disliking, similar to the in-person testing.








Research results and discussion:

Data analysis is currently being done on Phase I data. At this time, we do not have any findings to report. Data from Phase II is currently being collected.

Research conclusions:

Data analysis is currently being done on Phase I data. At this time, we do not have any findings to report. Data from Phase II is currently being collected.

Participation Summary

Education & Outreach Activities and Participation Summary

1 Webinars / talks / presentations

Participation Summary:

40 Number of agricultural educator or service providers reached through education and outreach activities
Education/outreach description:

Education and Outreach

In December 2019, graduate student Regina O’Brien presented an Educational Micro-Session at the New York Produce Show and Conference about her ongoing research on baby leafy greens. The NY Produce Show, co-sponsored by the Eastern Produce Council, is the second largest produce show in North America, attracting roughly 5,000 attendees and 400 exhibitors from the produce industry.

 The talk, “Controlled Environment Agriculture: A Tool to Understand Flavor Profiles and Consumer Demand for Baby Leafy Greens,” detailed the current collaborative research between AeroFarms (Newark, NJ) and the Center for Sensory Sciences and Innovation at Rutgers University-New Brunswick. The presentation highlighted the use of Controlled Environment Agriculture in Sensory Science and Food Science research and connected this with our plans to research consumer preference on baby leafy greens. In this talk, we were able to inform those in the produce industry about our research and acknowledged FFAR and NESARE for the ability to complete this research.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.