Please note, this is recovered content from the former website of the New York Obesity Research Center website.


Importance of location – Where You Eat Can Be Important To Acheive Real Insight About Eating Behaviour

Date: December 11, 2003
Speaker’s Name and Affiliation: Herbert Meiselman, U.S. Army Natick Labs
Topic: “Contextual Factors in Eating–Where You Eat is Important”

Eating SettingsDr. Meiselman presented an enlightening seminar to suggest that where you eat might be just as if not more important than what you eat. Quoting from his article in Appetite, 1992, 19, 49-55, researchers need to “refocus human eating research towards greater use of real meals, served to real people (not subjects), in real eating situations.” However, Dr. Meiselman continued, “this is not to suggest that all studies should be done in natural settings.”

Context according to the speaker is synonymous with ” location/setting, situation, environment.” Dr. Meiselman became interested in studying context during the mid 80s through the extended testing of army rations at US Army Natick Labs. When acceptance and consumption of rations were tested daily, it was noticed that although acceptance ratings for the foods were relatively high, consumption was relatively low.

This seems to contrast with other data (Guy-Grand et al 1994, Yeomans et al 1997, Bellisle & LeMangen 1980, McKenna 1972, Decke 1971 etc.) that have found that people eat larger meals when they are eating food they like. Several reasons for underconsumption of army rations were posited, including monotony of diet, quality, stress, disturbed sleep pattern, heavy energy expenditure, and novel food neophobia.

Talking Moose
Talking Moose
People eat larger meals when they are eating foods they like.

Dr. Halls Dr. Halls
Belongs in the Scientific Journal of “Captain Obvious”


Dr. Meiselman reviewed a wealth of other data from both his lab and from collaboration with other labs that suggest it is necessary to take context into account when measuring eating in or out of the lab.

In one experiment, a dish of fettuccine alfredo with chicken was served in the laboratory, dining hall, and training restaurant. When the dish was served in the restaurant, people rated gave higher acceptance ratings.

In the dining hall, the same dish was given lower acceptance scores, and in the laboratory intermediate scores. In another experiment that looked at the effect of enhanced setting on acceptance of different chardonnay products, context effects were as large as product effects.

David David
People eat more, in a fancy restaurant

Holly Holly
Except for girls, when on a date.


In concluding the extended work done with army rations, Dr. Meiselman summarized the following three points. 1) Acceptance does not always predict consumption. 2) Consumption may be a better indicator of acceptance in actual long term eating situations. 3) Situational variables must be considered to understand food choice, acceptance and consumption.

The speaker suggested that in the future, it will be a challenge to identify the 100s of factors that affect eating behavior, and how these factors are related to each other. Finally, Dr. Meiselman ended with the following intriguing statements. “You can study ingestion in the lab, but not eating.

You can study sensory in the lab, but not food appreciation. You must get out of the lab to validate your lab findings in real situations, and if necessary, to change your conclusions. You might even learn more about eating, and more about how to conduct your research.”


Q. What about your subject selection process?
A. We can’t choose subjects for our experiments. In our field tests, an intact military unit ( a platoon or company) is provided.

Q. Have you done experiments where you switch expectations, give people a typically good tasting food that actually tastes bad?
A. This type of work has been reported by Cardello on assimilation of expectations.

Q. In studying the differences in ratings between foods served in the refectory and the grill room, were foods priced the same or differently?
A. They were priced according to locale; restaurant foods were higher than cafeteria foods.

Q. Comment: It is interesting research, given that product developers base their decisions based on laboratory acceptance ratings.

Q. I am assuming that you don’t bring people from one setting to another in your experiments.
A. No, that is one limitation of location research. In the Edwards et al study in ten different locations we did observe the same population group in different locations, and we also had our first opportunity to get some demographic information. But generally, people and location are confounded as they are in real life.

Q. Do you see age differences in your work?
A. We find that younger people are more critical, but we are not sure at this point whether this is a people effect (younger people are more critical) or a location effect (more critical responses are found in locations which young people inhabit).

Q. How old are the patients in the elderly residential home?
A. 70-80

Q. If everything is between 6-8, why hasn’t the scale been changed?
A. The nine point hedonic scale is the industry standard, and everyone has large data bases using it, so it is difficult to change. But based on Bartoshuk’s last lecture on category scaling, we should all be examining scale use. Cardello at Natick has been working on some new labeled magnitude scales.

Q. Are you saying there is an interaction between food and context?
A. That is what the evidence shows; the same food in different locations is viewed differently.

Q. How do you deal with comfort foods?
A. I wouldn’t think a comfort food would be context independent.

Q. Is the increase in rating due to the fact that they are working all day (in army rations research)?
A. That may be a factor.

Q. This is similar to data where if you deprive people, they rate foods with higher acceptance scores.
A. Yes, that may be a factor, but these soldiers are not really deprived.

Q. Isn’t there evidence that if a food tastes better, people don’t need to eat as much to get the same effect?
A. I don’t think there is.

Q. Are gender differences in data from (Meiselman et al, 2003, slide 31) related to body size?
A. We didn’t get that.

Q. Do the soldiers gain weight if they go on leave?
A. I have no information on that.

Q. Why would they chose to lose weight in the field?
A. To maintain their weight standard.

Q. If they are purposely choosing their food for some reason (limited), isn’t it a specialized population (similar to restrained eaters)?
A. (Answer wasn’t recorded)

Q. Do you consider choice part of context?
A. Yes, choice is one of the variables which appears to differentiate different eating environments; laboratories usually have no choice, whereas many natural eating environments usually do.

Q. What do you think is the difference in using acceptability ratings of foods given alone vs. within a meal?
A. In a meal, ratings of the foods will be higher. The recent work with McCormick presented at Pangborn shows this clearly.

Q. This is interesting research. Is there any theoretical or general framework for context?
A. I am not aware of any theory to encompass context.

Comment: Several years ago, Nutrition Reviews published an article called “Determinants of Food Choice.” In that article, we have set up a model that might begin to generalize this research area.

Q. Have you noticed differences in the military since it has been volunteer vs. draft?
A. We have not examined this longitudinally, but studies of the early volunteer Army showed that food was one of he biggest disappointments of young soldiers.

Q. Do you include levels of hunger?
A. Yes, we sometimes ask soldiers to rate hunger, and we also measure the satiety value of the rations.