What’s for lunch? Eliciting preferences for food on university campus: discrete choice experiment protocol

doi:10.21203/rs.3.rs-4436883/v1

Download PDF

Study protocol

What’s for lunch? Eliciting preferences for food on university campus: discrete choice experiment protocol

https://doi.org/10.21203/rs.3.rs-4436883/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background. Food choices are influenced by habits, experiences, as well as various socioeconomic factors. Understanding these drivers can mitigate negative effects of poor nutrition and yield societal benefits. Preference elicitation methods like discrete choice experiments help understand people’s food preferences revealing factors influencing choices the most, such as nutritional content or cost of a meal. This information can be helpful in developing tailored meal-based interventions and informing food policies. Universities, as anchor institutions, are increasingly concerned with health, wellbeing, and sustainability of their students and staff. Yet, there is limited evidence on food preferences in university settings. This paper outlines a discrete choice experiment protocol to compare lunch preferences among university staff and students across six European countries, aiming to inform campus food policies.

Methods. Attributes and levels were derived from a systematic literature review of preference-based studies focused on the drivers of meal choices and validated in the focus group with students and staff from participating universities. The attributes in the discrete choice experiment include nutritional content, price, time to access a meal, sensory properties of a meal, naturalness of the ingredients, and meal size. The survey was piloted in think-aloud interviews with students and staff in participating universities. We will collect preference data, along with data on participants’ sociodemographic characteristics, food-related behaviour, opinions about food, experience of food insecurity, physical activity, and body composition, using an online survey. Preference data will be analysed using random parameter logit and latent class models.

Discussion. This study will be the first to investigate lunch preferences of university students and staff across six European countries, informing campus food policies. While campus food systems may not always align with students’ and staff preferences, incorporating them into policy-making can enhance satisfaction and well-being. Strengths include an international focus, inclusion of complementary variables, and involvement of potential respondents in all phases of developing this research. Acknowledging limitations, such as varying lunch habits, the study aims to provide valuable insights for improving university food policies and overall community well-being.

discrete choice experiment

food preferences

university

students

workplace

Every day, people make decisions about purchasing and consuming food. These decisions are guided by habits, past experiences, as well as a wide range of socioeconomic and cultural factors [1]. Food consumption has a direct impact on people’s health. Poor nutrition is associated with negative health (e.g. obesity, hypertension, cancer) (Schlesinger et al, 2019; Schwingshackl et al, 2017; Johnson et al, 2013), and[2–4], economic and environmental consequences (e.g. lower educational attainment, reduced workplace productivity, and greenhouse gas emissions) [5–7]. Having a better understanding of the drivers of food consumption choices can help mitigate some of these negative consequences and lead to health and economic benefits for individuals and society.

Preference elicitation methods are designed to help understand core preferences. There are several approaches to eliciting preferences, including discrete choice experiments (DCEs), that are gaining popularity in food research [8]. A DCE is a preference elicitation method that identifies the product attributes that influence people’s choices. For example, Livingstone and colleagues (2021) conducted a DCE to understand the relative importance of factors influencing young adults’ food choices [9]. They found that nutrition content and cost of a meal were the two most important influences of food choices, while preparation time was the least important. Preferences also varied by demographic and health characteristics. The authors concluded that a DCE is a suitable method for understanding the complexity of food choice behaviours and can generate useful evidence to support the design of tailored meal-based interventions.

While the majority of DCEs in food research have targeted the general population, several studies have been conducted to elicit people’s preferences for various attributes of meals in specific contexts such as schools [10, 11] and prisons [12]. Such evidence can be helpful in providing more nuanced understanding of food choice drivers in specific settings to inform local or organizational policies. For example, Rusmevichientong et al. (2021) used a DCE to quantify the relative importance of snack attributes among middle-school children suggesting that price and whole grains were the most important attributes for the choice of snack among the respondents [10].

Universities represent a unique setting as they include diverse populations in terms of age, ethnicity, and cultural backgrounds. With being both a place of education and a workplace, universities act as ‘anchor institutions’ within a food system with responsibility for providing a healthy environment for their staff and students. Universities are becoming increasingly interested in the health and wellbeing of their students and staff as well as in becoming more environmentally sustainable. This is evidenced by, for example, The UK Healthy Universities Network aiming to support universities with developing and implementing ‘whole university’ approaches to health, wellbeing, and sustainability [13], and the work of The Association for the Advancement of Sustainability in Higher Education (AASHE) with empowering higher education staff and students to drive sustainability innovation [14], among other initiatives. However, there is a paucity of evidence on preferences for food provision in university settings, and understanding the drivers of students’ and staff preferences for food offered on campus will provide valuable evidence for informing university food policy.

This paper presents the protocol for conducting a DCE to elicit and compare preferences of university staff and students for lunch, as lunch is the most commonly bought meal on campus by both students and staff, across different country settings. The study will be conducted within six university campuses located within six European countries (France, Hungary, Italy, Spain, Sweden, and the United Kingdom).

Context

This study will be conducted in universities participating in the EUniWell alliance [15], including University of Birmingham (The United Kingdom), University of Florence (Italy), Linnaeus University (Sweden), Nantes University (France), Semmelweis University (Hungary), and University of Murcia (Spain). Characteristics of each university are presented below (Table 1).

Table 1

University characteristics
	University of Birmingham*	University of Florence	Linnaeus University	Nantes university	Semmelweis University	University of Murcia
Number of students	36,933	53,612	33,000	43,000	14,024	31,015
Proportion of international students	31%	8%	2%	12%	35%	3%
Number of staff	9,053	4,399	2,100	4,500	2,095	3,839
Proportion of international staff	24%	Unknown	Unknown	Unknown	Unknown	Unknown
Location	Urban	Urban and surroundings	Urban and surroundings	Urban and surroundings	Urban	Urban and surroundings
Number of campuses	3 (Edgbaston, Selly Oak, and the Dubai campus overseas)	7	2 (Växjö and Kalmar)	7 (Nantes, Carquefou and 2 surrrounding cities, Saint-Nazaire, Roche/Yon)	The university does not have a campus. Instead, its faculties, departments, hospitals, clinics, libraries, sport and accommodation facilities are scattered throughout the capital city of Budapest	5
Type of campus (city or campus-based)	Campus-based	Both city and campus-based	Campus-based	City and campus-based	City-based	Campus-based
*all estimates exclude the students and the staff at the Dubai campus of the University of Birmingham

Study design: Discrete choice experiment (DCE)

A DCE is a method for eliciting preferences for the attributes of a product or service. Using survey methodology, DCEs assess preferences by asking individuals to make choices or trade-offs between hypothetical options that differ according to their attributes. This method is based on the principle that people derive utility, i.e., wellbeing, for a product/service from its attributes, and that the choices revealed through a DCE enable inferences on the relative contribution of each attribute and level to the overall utility of a product or a service. DCEs are based on strong theoretical underpinning (random utility theory) and are currently seen as the gold standard for evaluating preferences [16]. The development of a DCE involves a series of steps including the identification of attributes and levels, experimental and instrument design, data collection, and statistical analysis [17].

The overview of the study steps to develop the DCE is presented in Fig. 2 below, including a systematic literature review to identify the initial list of attributes and levels, focus group discussion using the nominal group technique approach to validate the list of attributes and levels, and think-aloud interviews to pilot test the survey.

Figure 1. Steps to develop the discrete choice experiment survey

Identification of attributes and levels

To identify potentially relevant attributes and attribute levels, we conducted a systematic literature review of preference-based studies focused on the drivers of meal choices [18]. The objectives of the review were to summarise the evidence generated from DCEs and other preference-based methods to understand meal-choice; and to identify a list of attributes for the development of a DCE to investigate demand for lunch on campus. We constructed a comprehensive search strategy and searched Web of Science, Scopus, Medline, Embase, PsychINFO, EconLit, and CINAHL to identify eligible studies. After title/abstract and full-text screening, 33 studies were included in the review. The important constructs, in terms of attributes and their levels, were extracted from the identified studies. They were clustered into groups corresponding to similar themes (e.g. taste, price), and ranked in terms of frequency of being included in the identified DCEs as well as their relevance to the project setting. The clarity and mutual exclusivity of the attributes on the list were further discussed by the project team. This resulted in the initial list of nine potentially relevant attributes accompanied by descriptions and corresponding levels (Table 2).

Table 2

List of attributes, descriptions and levels generated from the literature
	Attribute	Description	Levels
1	Environmental impact	Negative impact of food production on the environment (e.g. CO2 emissions, water usage, animal welfare)	Low; moderate; high
2	Food origin	Origin of the ingredients	Locally sourced; imported; unknown
3	Healthiness	Meal that is good for your health	Healthy; neutral; unhealthy
4	Price	Price paid for food	Cheap; average; expensive
5	Time	How fast the food is to access (including walking to the food outlet and waiting for the meal to be prepared)	Fast; moderate; slow
6	Sensory properties of a meal	Appearance, smell, taste, texture, and colour of the meal	OK; good; very good
7	Familiarity	Familiarity with the food options available	Not very familiar; somewhat familiar; very familiar
8	Naturalness	The extent of processing during the food production process	Minimally processed; processed; ultra-processed
9	Serving size	Size of the portion	Small; average; big

Validation of the list of attributes and levels

From this initial list, the final list of attributes accompanied by descriptions and levels were determined using the nominal group method in a focus group. This technique has frequently been used to select attributes for DCEs [19]. The original aim was to recruit one member of staff and one student from each participating university, however due to difficulties with recruitment and delays with obtaining ethics approvals, the final sample for the nominal group included eight participants from five out of the six participating universities.

The participants were presented the initial list displayed in Table 2 and asked to:

Discuss the clarity and mutual exclusivity of the attributes.

Discuss the clarity of the descriptions.

Prioritise the attributes in terms of their importance when selecting a meal.

Discuss the appropriateness of the levels.

The session was conducted online and led by a moderator (IP). During the session, the participants were asked to think about choosing what to eat for lunch on a typical day on campus, when they did not bring their own lunch. During the session, participants discussed the initial list of attributes presented in Table 2; as a result of this discussion some attributes were combined (e.g. healthiness and naturalness) and new attributes were introduced (e.g. variety of meal options available and environment in which a meal is consumed). Participants were asked to provide written consent prior to participation and received shopping vouchers (value of £25/€30)

In accordance with good practices, we aimed to include a maximum of 6 attributes in the final list to avoid overburdening the DCE respondents [20]. Therefore, the focus group participants were asked to select the top-6 attributes they found most important when choosing lunch on campus. The results of the voting are presented in Table 3 below.

Table 3

Results of the attribute prioritization
Order of importance based on individual voting	Attribute
1st	Price of a meal
2nd	Nutritional content of a meal
3rd	Time it takes to walk to the outlet and wait for a meal to be served
4th	Sensory properties of a meal
5th	Variety of meal options available
6th	Naturalness of the ingredients used to prepare a meal
7th	Serving size of a meal
8th	Environment in which a meal is consumed
9th	Familiarity with a meal
10th	Environmental impact of a meal

Based on the focus group discussion and voting, and subsequent discussion amongst the project team, the final list of attributes, descriptions, and levels were developed (Table 4). It is important to note that even though the focus group participants placed variety of meal options in the top-6, it was excluded from the final list of attributes, because variety was an attribute describing a collection of meal offers, rather than the attributes of a single meal.

Table 4

Final list of attributes, descriptions and levels
Factor	Description	Level
Nutritional content	How well a meal is able to meet your nutritional and dietary needs	Insufficient; neutral; sufficient
Price	Price paid for a meal	20% below average; average; 20% above average
Time	How fast the food is to access (including walking to the food outlet and waiting for the meal to be prepared)	Fast; moderate; slow
Sensory properties of a meal	Appearance, smell, taste, texture, and colour of a meal	Poor; OK; very good
Naturalness	The extent of processing of the ingredients during the food production process	Minimally processed; processed; ultra-processed
Meal size	Amount of food you receive in a meal	Small; medium; large

Design of choice tasks

The DCE choice tasks were designed based on the final list of attributes and corresponding levels displayed in Table 3 using dcreate package in Stata version 17.0 (StataCorp LLC, College Station, TX). This package allows for the development of an efficient design that maximizes the precision of estimates by using a-priori information on the levels for each attribute that were generated from the nominal group discussion. In line with good practice, we generated 3 blocks of 8 choice tasks. Furthermore, we included a test-retest validity question, a question that repeats twice in each choice, to test the consistency of respondents’ preferences. The test-retest validity questions are there as a validation check as if any respondent ‘fails’, they are excluded from further analysis. Within the DCE, each respondent will be randomly allocated to one of the three blocks of choice tasks. Choice scenarios will be presented using visual aids to ease comprehension. Figure 2 below shows an example choice task.

Figure 2. Example of a choice set in the discrete choice experiment

Survey design

The DCE will be incorporated within a larger survey that will also include questions about participants’ sociodemographic characteristics, food-related behaviour (e.g. typical source of lunch, usual diet, food allergies), opinions about food, experience of food insecurity, physical activity, and body composition. We will also assess the level of burden of survey completion. The original survey will be developed in the English language and translated into other languages (French, Swedish, Italian, Hungarian, and Spanish) by an external translation agency. Translations will be checked for correctness by the researchers (native speakers) in each participating university.

Pilot test of the DCE survey

Earlier versions of the survey were pilot tested with seven staff members and six students from the participating universities using a think-aloud interview approach. During these interviews we assessed the comprehensibility of the survey and the difficulty of completing it. Overall, participants found the survey understandable albeit rather lengthy and suggested some clarifications, for example, adding definitions to the listed diet types and eating patterns. It took them on average 15–20 minutes to complete. Based on the feedback from the pilot, we added clarifications and reduced the number of choice tasks from 12 to 8. All pilot participants signed consent forms and received a shopping voucher (value of £25/€30). The final version of the DCE survey is included in Supplementary File 1.

Data collection and sampling

To conduct the DCE survey, we will use Qualtrics^XM (Qualtrics, Provo, UT) for the respondents from all participating universities except for the Linnaeus University, for which the survey will be developed using Survey&Report (Artisan). The invitation to participate will be distributed to students and staff from the six participating universities using various media (e.g. university newsletters, flyers, directed emails, social media promotion, etc). In our study, we will aim to recruit at least 100 respondents from each category, i.e. 100 staff members and 100 students, from each participating university. This was sufficient according to the common rule-of-thumb estimation for DCEs [21].

Statistical analysis

The participant characteristics will be described using Stata software, version 17.0 [22]. Data on preferences will be analysed using Nlogit 6 [23]. Since all attribute levels are categorical, they will be coded using the effects-coding approach. For this analysis, one level of each attribute is omitted and non-omitted variables are assigned a value of 1 when they are present, and 0 when another non-omitted variable is present. In effects coding, non-omitted variables are assigned the value of -1, when an omitted variable is present. Effects coding yields a unique coefficient for each attribute level included in the study.

We will first analyse the choice data using a random parameter logit model that will capture preference heterogeneity. We will also estimate subgroup random parameter logit models to assess if the preferences varied as a function of participant characteristics (sociodemographic characteristics and health-related behaviours). Finally, we will estimate a latent class model to identify preference classes based on participants’ preferences. All participants who completed the DCE-part of the survey and passed the test-retest validity check will be included in the main analysis. Preference data from all participants regardless of the test-retest validity check will be analysed separately in a secondary analysis.

Ethics

Prior to their participation in the survey, all participants will be provided with clear information on the study aims and objectives and asked to provide consent. Participants will have the option to drop out of the survey at any point without providing a reason or facing consequences. Ethical approval for this study was sought from the university ethics committee in each participating university. This study was approved by the ethics committees of the University of Birmingham (ERN_1270-Jun2023), University of Murcia (M10/2023/046), Nantes University (n°031020230), University of Florence (n. 304 granted on 21/02/2024), and Swedish Ethical Review Authority (Dnr 2023-07604-01).

To the best of our knowledge, this will be the first study to elicit preferences of university students and staff for lunch on campus in six universities in six European countries contributing evidence to inform university food system policies. Food systems, including those on university campuses, are influenced by a multitude of factors including business considerations and statutory regulations and may not always be in line with what the customers, i.e. students and staff members, prefer, or offer food choices that maximise health and environmental outcomes. Incorporating their preferences in how food policies are developed can offer a new perspective to decision-makers and help enhance the satisfaction and well-being of university students and staff.

Studies adopting a DCE design require a well-defined research question and scope for specifying the context for making choices [17]. Therefore, our DCE will focus on a lunch-time meal bought on campus. Our findings will likely to be less applicable to other eating occasions, such as breakfast or snack, but given that the majority of students and staff spend time on campus during day time, we can reasonably assume that the largest proportion of food consumption on campus occurs at lunch.

Attributes pertaining to the overall experience of out-of-home meal consumption, such as food offer (e.g. variety of meal options), environment in which meals are consumed (e.g. whether food is served by waiters or self-serviced), or the social element of eating were not included in our DCE, even though such attributes have been found to be important drivers of meal choice in previous preference-based studies [24, 25]. Specific focus on meal attributes will enable us to examine and compare preferences for those attributes that are universally applicable to meals regardless of the country and university context. At the same time, the relative importance of the included attributes will offer valuable insights that will inform university food policy.

A particular strength of this study will be its international focus given that the proposed survey will be simultaneously conducted in six universities across six European countries. This research will help to fill the gap in the literature on the determinants of demand for food on campus and potential differences between various countries. We followed good practice guidelines for development of DCEs in the healthcare setting [17], including conducting a comprehensive literature review to identify the initial list of attributes and levels [18], a focus group discussion, and pilot think-aloud interviews with university students and staff from the participating universities. This will ensure the DCE is applicable to the different country settings. This is an important strength of this study given the underreporting of this step in previous DCE research [8]. The results will be reported through various means including academic conferences and a peer-reviewed publication to communicate the findings to the academic community, and through reports and policy briefs to reach university policymakers.

Another strength of the study is through the deliberate inclusion of complementary variables aimed at improving our understanding of preference formation. First, the identification of relationships between these preferences and specific behaviours, such as the source of lunch, dietary habits, and food allergies and intolerances, is in line with the principles of behavioural economics [26]. Through the incorporation of these variables, we will provide a more nuanced understanding of the determinants of food choices among university staff and students. Moreover, because we consider food involvement, i.e., the importance of food for the subject, we will add a layer of depth to the analysis, capturing a motivational dimension that explains the extent to which subjects engage in information processing during food decision-making processes.

It is important to acknowledge that a proportion of university students and staff typically bring their own lunch to campus, and this proportion may vary across country settings. However, we assume that the majority of respondents would still buy lunch on campus from time to time. While we will not be able to explore this in-depth, in our survey, we will account for the average number of days spent on campus by the participants and their typical source of lunch.

The heterogeneity of food systems across the different campuses, as well as any underlying cultural differences, might present challenges to the conduct of the DCE. We plan to mitigate this by incorporating the local context knowledge by working closely with research partners from each university, and with involving potential respondents in the design and pilot stages, developing translated versions of the survey, and accounting for heterogeneity in the analysis through the inclusion of a wide range of explanatory co-variates.

In summary, by taking into consideration the interrelationships of food preferences with related behaviours and food involvement, our study will provide valuable insights for practical applications within the university setting, with the ultimate goal of contributing to food policy that will lead to a positive impact on the university community well-being.

Ethics approval and consent to participate

This study was approved by the ethics committees of the University of Birmingham (ERN_1270-Jun2023), University of Murcia (M10/2023/046), Nantes University (n°031020230), University of Florence (n. 304 granted on 21/02/2024), and Swedish Ethical Review Authority (Dnr 2023-07604-01). Focus group and pilot interview participants signed informed consent forms prior to participation.

Consent for publication

This manuscript does not contain any personal data.

Availability of data and materials

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

Competing interests

The authors declare that they have no competing interests

Funding

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101035821.

Authors' contributions

All authors (IP, NA, LF, MH, CW, MH, LDN, FR, AJE, KB, ÉCC, KO, PP, SRDM, EFB, and EF) made substantial contributions to securing funding for this work, to the conception and design of the work. IP, NA, CW, MH, LDN, FR, AJE, KB, ÉCC, KO, PP, SRDM, EFB, and EF contributed to the acquisition, analysis, and interpretation of the focus group and pilot interview data. MH provided technical expertise on the analysis of the discrete choice experiment data.

IP drafted the manuscript. All authors contributed to revising the manuscript. All authors approved the submitted version of the manuscript and agreed both to be personally accountable for the author's own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which the author was not personally involved, are appropriately investigated, resolved, and the resolution documented in the literature.

Acknowledgements

The authors would like to thank focus group and pilot interview participants for their time and help in developing the survey.

Neumark-Sztainer D, Story M, Perry C, Casey MA. Factors influencing food choices of adolescents: findings from focus-group discussions with adolescents. J Am Diet Assoc. 1999;99(8):929–37.
Schlesinger S, Neuenschwander M, Schwedhelm C, Hoffmann G, Bechthold A, Boeing H, et al. Food groups and risk of overweight, obesity, and weight gain: a systematic review and dose-response meta-analysis of prospective studies. Adv Nutr. 2019;10(2):205–18.
Schwingshackl L, Schwedhelm C, Hoffmann G, Knüppel S, Iqbal K, Andriolo V, et al. Food groups and risk of hypertension: a systematic review and dose-response meta-analysis of prospective studies. Adv Nutr. 2017;8(6):793–803.
Johnson CM, Wei C, Ensor JE, Smolenski DJ, Amos CI, Levin B, et al. Meta-analyses of colorectal cancer risk factors. Cancer Causes Control. 2013;24:1207–22.
Fard NA, Morales GDF, Mejova Y, Schifanella R. On the interplay between educational attainment and nutrition: a spatially-aware perspective. EPJ Data Sci. 2021;10(1):18.
Cho S, Kim S. Does a healthy lifestyle matter? A daily diary study of unhealthy eating at home and behavioral outcomes at work. J Appl Psychol. 2022;107(1):23.
Grosso G, Fresán U, Bes-Rastrollo M, Marventano S, Galvano F. Environmental impact of dietary choices: role of the Mediterranean and other dietary patterns in an Italian cohort. Int J Environ Res Public Health. 2020;17(5):1468.
Lizin S, Rousseau S, Kessels R, Meulders M, Pepermans G, Speelman S, et al. The state of the art of discrete choice experiments in food research. Food Qual Prefer. 2022;102:104678.
Livingstone KM, Abbott G, Lamb KE, Dullaghan K, Worsley T, McNaughton SA. Understanding meal choices in young adults and interactions with demographics, diet quality, and health behaviors: A discrete choice experiment. J Nutr. 2021;151(8):2361–71.
Rusmevichientong P, Jaynes J, Chandler L. Understanding influencing attributes of adolescent snack choices: Evidence from a discrete choice experiment. Food Qual Prefer. 2021;92:104171.
Hartmann M, Cash SB, Yeh C-H, Landwehr SC, McAlister AR. Children's purchase behavior in the snack market: can branding or lower prices motivate healthier choices? Appetite. 2017;117:247–54.
de Guzman A, Barredo SF, Caillan KR. Examining the role of depression in the Filipino elderly’s food preferences in prison setting: Data from conjoint analysis and SEM. Int J Prison Health. 2020;16(2):135–49.
The UK Healthy Universities Network. [ https://healthyuniversities.ac.uk/.
The Association for. the Advancement of Sustainability in Higher Education [ https://www.aashe.org/.
European University for Well-Being. 2024 [ https://www.euniwell.eu/.
Whitty JA, de Bekker-Grob EW, Cook NS, Terris-Prestholt F, Drummond M, Falchetto R, et al. Patient preferences in the medical product lifecycle. Patient-Patient-Centered Outcomes Res. 2020;13(1):7–10.
Bridges JF, Hauber AB, Marshall D, Lloyd A, Prosser LA, Regier DA, et al. Conjoint analysis applications in health—a checklist: a report of the ISPOR Good Research Practices for Conjoint Analysis Task Force. Value health. 2011;14(4):403–13.
Afentou NFL, Frew E, Pokhilenko I. Systematic literature review of discrete choice experiments of meal preferences. [Article under review.]. In press. 2024.
Hiligsmann M, van Durme C, Geusens P, Dellaert BG, Dirksen CD, van der Weijden T et al. Nominal group technique to select attributes for discrete choice experiments: an example for drug treatment choice in osteoporosis. Patient Prefer Adherence. 2013:133–9.
de Bekker-Grob EW, Ryan M, Gerard K. Discrete choice experiments in health economics: a review of the literature. Health Econ. 2012;21(2):145–72.
de Bekker-Grob EW, Donkers B, Jonker MF, Stolk EA. Sample size requirements for discrete-choice experiments in healthcare: a practical guide. Patient-Patient-Centered Outcomes Res. 2015;8:373–84.
StataCorp. Stata Statistical Software: Release 17. College Station. TX: StataCorp LLC; 2023.
Nlogit. Superior Statistical Analysis Software 2024 [ https://www.limdep.com/products/nlogit/.
Ernawati H, Suwandojo DPEH. Consumer preferences for indonesian food. J Indonesian Econ Business: JIEB. 2019;34(3):280–93.
Landwehr SC, Hartmann M. Is it all due to peers? The influence of peers on children's snack purchase decisions. Appetite. 2024;192:107111.
Lancsar E, Louviere J. Conducting discrete choice experiments to inform healthcare decision making: a user’s guide. PharmacoEconomics. 2008;26:661–77.

No competing interests reported.

SupplementaryFile1.docx

Download PDF

Submission checks completed at journal
22 May, 2024
Editor assigned by journal
22 May, 2024
First submitted to journal
17 May, 2024

You are reading this latest preprint version

What’s for lunch? Eliciting preferences for food on university campus: discrete choice experiment protocol

Status:

Version 1

Abstract

Figures

Introduction

Methods

Context

Study design: Discrete choice experiment (DCE)

Identification of attributes and levels

Validation of the list of attributes and levels

Design of choice tasks

Survey design

Pilot test of the DCE survey

Data collection and sampling

Statistical analysis

Ethics

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1