Food Quality and Preference 18 (2007) 58–65

www.elsevier.com/locate/foodqual

The effects of contextual variables on food acceptability:

A confirmatory study
a,*
Silvia C. King , Herbert L. Meiselman b, Annette W. Hottenstein a,
Therese M. Work a, Valerie Cronk c
a
McCormick and Company, 204 Wight Avenue, Hunt Valley, MD 21030, USA
b
US Army Natick Soldier Center, Natick, MA 01760–5020, USA
c
Olive Garden, 5900 Lake Ellenor Drive, Orlando, FL 32809, USA

Received 25 February 2005; received in revised form 11 July 2005; accepted 31 July 2005
Available online 21 September 2005

Abstract

Consumer acceptance of food and beverage was measured in three different settings/locations: a central location test in a labo-
ratory facility, a central location test at one unit of a national chain restaurant and a customer satisfaction survey at the same chain
restaurant in multiple cities. Two main dishes (lasagna, cannelloni), salad, breadsticks and iced tea were served either as individual
items or as part of a meal. Meal context and the consumerÕs ability to choose had the strongest positive effects on acceptance ratings,
while social interaction and enhanced environment had no noticeable effects on the acceptability scores. There were significant age
and gender effects in the two restaurant settings, but not in the laboratory central location test. The results of this study confirm
some of the results of King et al. (2004) [King, S., Weber, A., Meiselman H., & Lv, N. (2004). The effect of meal situation, social
interaction, physical environment and choice on food acceptability. Food Quality and Preference, 15, 645–653] on the enhancing
effects of context variables on product acceptance, and on how the relationship between context effect and consumer acceptance
may not be consistent within and across meal components.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: Acceptability; Consumer; Context effect; Choice; Environment; Meal; Menu items; Restaurant; Sensory; Social interaction

1. Introduction Rozin (1996) described the range of contexts that can

influence eating and food choice, including indirect envi-
One of the goals of a product test with consumers is ronmental/cultural factors, indirect personal factors and
to increase product acceptance in the actual market- socio-cultural aspects which occur prior, during and
place, the real world. Context has been identified as an after food consumption. Tuorila, Meiselman, Bell, Car-
important factor in understanding the acceptance of dello, and Johnson (1994) demonstrated that setting
food and in predicting food consumption (Meiselman, expectations as part of the context in the evaluation of
Hirsh, & Popper, 1988). Rozin and Tuorila (1993) and a novel food can help improve the acceptability of the
product. Meiselman (2002) identified four of the context
factors which can influence food acceptability: evaluat-
ing the food as part of a meal, the effect of social inter-
*
Corresponding author. Tel.: +1 410 771 7390; fax: +1 410 527
action during food consumption, the environment in
6527. which the food is consumed, and the ability to make
E-mail address: [email protected] (S.C. King). choices regarding the food that is being consumed.

0950-3293/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodqual.2005.07.014
 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65 59

King, Weber, Meiselman, and Lv (2004) demonstrated 2. Materials and methods

that these context factors have an enhancing effect on
food acceptability when comparing traditional labora- 2.1. Approach
tory tests with meals served in modified laboratory or
real restaurant settings. Serving food as part of a meal Three tests were conducted to compare a laboratory
and offering a choice of foods have the greatest positive test to restaurant settings for consumer responses to a
impact on acceptability, followed by enhancement of the meal composed of iced tea, breadsticks, salad, lasagna
testing environment. Meal components have been and cannelloni. These three approaches incorporate
shown to have different weights and impact on the over- one or more of the following contextual factors: evaluat-
all acceptability of a food. Eindhoven and Peryam ing the food as part of a meal, allowing social interac-
(1959) demonstrated that the acceptability of a food tion, enhancing the testing environment and providing
changed when evaluated as a single item versus evalu- food choices (Table 1). Test 1 was conducted at the
ated as part of a meal. McCormick Sensory Science Center in Hunt Valley,
Meiselman, Johnson, Reeve, and Crouch (2000) and MD and consisted of a standard central location test
Edwards, Meiselman, Edwards, and Lesher (2003) dem- (CLT). In Test 2, samples were served at one unit of a
onstrated that differences in acceptability of the same national Italian chain restaurant. Test 3 was a nation-
product can be influenced by the eating environment. wide in-store satisfaction survey performed at other
There is a clear and repeatable difference between the units of the same Italian restaurant chain. In this test,
acceptability of institutional food and restaurant food. respondents only filled out a survey for one rather than
Bell, Meiselman, Pierson, and Reeve (1994) demon- all five meal items (tests 1 and 2). Tests 1 and 2 each
strated that manipulating the ethnicity of the environ- took one day in three 1-hour sessions during the late
ment influenced response to the same food. Identical morning, afternoon and evening (11:00, 15:00 and
foods were tested in a restaurant with different ethnic 17:30). Test 3 was a restaurant satisfaction survey that
decors demonstrating that when the environment rein- took place during the evening, over the course of several
forces the food theme, acceptability increases. Beatty weeks at over 40 locations across the United States.
(1982), Kramer, Lesher, and Meiselman (2001) investi-
gated the effect of choice on food acceptability. Beatty 2.2. Sample preparation/presentation
found that females responded positively to variety by
consuming more food, while malesÕ consumption re- Five items were served: unsweetened iced tea, tossed
mained unchanged. Kramer et al. (2001) found that peo- salad (mixed greens, red onion rings, sliced tomato, ol-
ple who select the same food repeatedly in a choice ives, peperoncini and croutons) with Italian dressing,
situation actually rate that food higher than people garlic breadsticks, cannelloni with meat filling and meat
who select the food less frequently––this argues against lasagna. Ingredients were prepared according to the res-
the rule that more variety is always better, and under- taurantÕs specification in all three tests. Items in test 1
scores the importance of considering choice. were prepared by Olive Garden and McCormick chefs.
The purpose of this study was to expand the results Test 2 items were prepared by the same chefs. The test
from an initial context study (King et al., 2004) by was executed at a local (Baltimore area) restaurant. Test
repeating some of the tests with different food items 3 items were prepared by the individual restaurant chefs.
and environments. King et al. studied a variety of en- The test was executed in 40 restaurants across the Uni-
hanced laboratory settings and one restaurant setting, ted States. Serving sizes and presentation order varied
a small local casual restaurant. In the present study by test as described in Table 2.
there is only one laboratory-based setting and two differ-
ent tests in a national chain restaurant setting. The over- 2.3. Ballot
all goal of this research is to identify better tools to
predict consumer acceptability of products in natural In tests 1 and 2, a 9-point structured hedonic scale
environments. (1 = dislike extremely; 5 = neither like nor dislike;

Table 1
Testing protocol
Context Test
Test 1: traditional CLT Test 2: restaurant CLT Test 3: restaurant survey
Meal Individual meal components Individual meal components Meal
Social Self Social Social
Environment Consumer testing facility Restaurant Restaurant
Choice No choice No choice Choice
This protocol contains the context factor options included in each test.
60 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65

9 = like extremely) was used for overall liking (Peryam the overall rating. Overall acceptability was placed as
& Pilgrim, 1957). To provide consistency in ballot design the first question in the ballot for tests 1 and 2 (test 3
across all three tests, a 6-point structured hedonic scale did not include an overall acceptability measure). Over-
(1 = poor; 2 = fair; 6 = excellent) was used to measure all rating was measured in all three tests and was imbed-

Table 2
Experimental conditions by test to investigate four context effects: meal, social, environment and choice
 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65 61

ded in the ballot among other liking and diagnostic Compensation was provided for tests 1 ($30) and 2
questions. Overall rating is the primary measure of ($50 restaurant gift card); no compensation was pro-
acceptability being presented in this paper. vided for test 3.
Additional diagnostic information was collected but
will not be discussed in this paper. Demographic infor- 2.5. Data analysis
mation (gender, age, and restaurant usage frequency)
was also collected at the end of each questionnaire. In Data were analyzed using univariate and multivariate
tests 1 and 2, the questionnaire containing a page with analysis of variance procedures of the SAS system
questions for each of the items evaluated, was presented (Cary, NC). Duncan means separation test was used
at the beginning of the test. In test 3, the ballot (satisfac- to differentiate samples. Differences described through-
tion survey) was presented with the meal item evaluated. out the paper refer to statistically significant differences.

2.4. Consumers and test methodology

3. Results
Participants for tests 1 (n = 74) and 2 (n = 83) were
recruited via McCormickÕs automated telephone recruit- 3.1. Overall test comparison
ing system, and were comprised of males and females,
ages 18–65. Individuals were required to be consumers There were significant differences (p < 0.0001) among
of Italian pasta dishes with meat sauces (lasagna and the tests when average ratings of tea, salad, breadsticks,
cannelloni were used as examples), garlic breadsticks, lasagna and cannelloni were combined (Table 3). Over-
salad and iced tea, to qualify for the test. Test 1 items all rating results are based on a 6-point hedonic scale
were served following a sequential monadic complete where the mean score for test 3 (5.0) was significantly
block design; social interaction was not permitted; the higher than test 1 (4.1) and test 2 (4.0). There was no dif-
room was set-up for individual evaluations, consumers ference in overall acceptability scores (9-point hedonic
were not given a choice of which product(s) to test, scale) between test 1 (7.2) and test 2 (7.1). No measure
and monetary compensation was provided at the end was available for test 3. Note that there were 2 different
of the test. measures of acceptability in this study: Overall
Three sessions were conducted during the in-restau- acceptability (9-point hedonic scale) and overall rat-
rant CLT (test 2), at 11:00, 15:00 and 17:30. In test 2, ing (6-point scale used by Olive Garden). Overall
items were presented in the order typically served at acceptability was measured in tests 1 and 2. Overall rat-
the restaurant; each item was evaluated individually ex- ing was measured in all three tests. Overall rating is the
cept for the beverage, which was available throughout primary measure of acceptability being presented in this
the test. It should be noted that consumers evaluated paper.
both entrees during the test; however the presentation
of these entrees was randomized among dinning tables. 3.2. Meal components
Consumers sat in groups of 4–6 per table to allow for so-
cial interaction. Results from individual meal components (salad,
Test 3 was conducted at various restaurant locations breadsticks, lasagna, cannelloni and tea) for each of
nationwide, with patrons of the restaurant who ordered the test conditions are shown in Table 4. Significant dif-
the specific item and completed the customer satisfac- ferences (p < 0.0001) existed among the tests for each
tion survey during their meal. The surveys were collected meal component except for lasagna, where scores re-
during normal business hours (lunch or dinner). The mained constant. Testing foods as items (tests 1 and 2)
survey was presented at the time the item was served rather than as part of a meal selected from the restau-
and consumers were asked to complete the question- rant menu (test 3) yielded lower scores for all items ex-
naire during the consumption of the item. Consumers cept lasagna. Test 2, which was the CLT conducted at
evaluated one item only. Data were collected over the the restaurant, yielded similar results to test 1, which
course of several weeks. The respondent base for test 3 was conducted at a testing facility. It should be noted
ranged from 250 to 390 for each of the five menu items. that the ranking order of the mean scores for the items

Table 3
Mean values for overall rating and overall acceptability for all items combined, across tests
Attribute Test 1: traditional CLT Test 2: restaurant CLT Test 3: restaurant survey P value
(n = 74) (n = 83) (n = 386)
Overall rating 4.1b 4.0b 5.0a <.0001
Overall acceptability 7.2 7.1 – 0.2565
CLT: central location test.
62 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65

Table 4
Mean values for overall rating for each menu item across tests
Meal component Test 1: traditional CLT Test 2: restaurant CLT Test 3: restaurant survey P value
(n = 74) (n = 83) (n = 386)
Lasagna 4.9 4.7 5.0 0.2823
Cannelloni 3.7b 3.5b 4.7a <.0001
Iced tea 3.8b 3.7b 4.9a <.0001
Salad 4.1b 4.2b 5.2a <.0001
Breadsticks 3.8b 3.7b 5.0a <.0001
One-way analysis of variance was used to compare the effect of test for each meal component. Within row, means sharing letters are not significantly
different. 6-Point overall rating score where 1 = poor, 2 = fair, . . . , 6 = excellent.
CLT: central location test.

Table 5
Mean values for overall rating for each menu item or all items (5.1 versus 4.6) and iced tea (4.6 versus 4.1) higher than
combined, across gender
males. These differences were driven primarily by the re-
Overall rating Males Females P value sults from test 3 (p < 0.0001). There were no significant
n = 287 n = 441 differences in the other items (cannelloni, salad and
Combined 4.5b 4.7a <.0001 bread).
n = 161 n = 276 There were significant differences (p < 0.0001) among
Lasagna 4.6b 5.1a <.0001
n = 154 n = 184 age groups on the overall rating for all the items com-
Cannelloni 4.3 4.2 0.4976 bined (Table 6). Participants aged 45–54 scored prod-
n = 133 n = 214 ucts lower overall for all the products combined as
Iced tea 4.1b 4.6a 0.0049 well as for individual products. Participants aged 18–
n = 140 n = 282
24 and 55 and higher rated the products higher in gen-
Salad 4.7 4.9 0.0684
n = 162 n = 277 eral, and participants aged 25–44 scores fell somewhere
Breadsticks 4.5 4.5 0.9141 between the other groups. There were no gender x age
interactions (p = 0.63).
Gender and age scores were compared within each
remained consistent across all tests. Lasagna consis- test. In test 1 (lab CLT), there were no significant differ-
tently ranked first, followed by the salad (second), tea ences between genders or ages. In test 2 (restaurant
and bread (tied for third place, except in test 3, where CLT), there were significant age x gender interactions,
tea ranked fourth), and cannelloni ranked last. as well as differences among age group responses. Fe-
males in the 18–24 age group scored the products
3.3. Gender and age effects lower than males (0.5 scale point difference), while the
females in the 25–45 groups rated the products higher
There were significant differences (p < 0.0001) be- (about 0.4 scale point difference). In test 3 (restaurant
tween genders on the overall rating scores for all the survey), females rated the products higher than males
items combined (Table 5). Females rated the products (5.1 versus 4.8) but differences among ages were not
higher (4.7) than males (4.5). Females rated lasagna significant.

Table 6
Mean values for overall rating for each menu item or all items combined, across age
Overall rating 18–24 yr 25–34 yr 35–44 yr 45–54 yr 55–64 yr 65+ yr P value
n = 287 n = 441 n = 464 n = 598 n = 151 n = 122
Combined 5.1a 4.7b 4.5b 4.3c 5.0a 5.0a <.0001
n = 71 n = 91 n = 104 n = 124 n = 33 n = 35
Lasagna 5.2 4.8 4.9 4.8 4.8 5.2 0.0878
n = 36 n = 81 n = 87 n = 107 n = 21 n = 21
Cannelloni 4.7ab 4.3bc 4.4bc 3.9c 5.1a 4.8ab 0.0001
n = 38 n = 81 n = 78 n = 114 n = 32 n = 15
Iced tea 5.1a 4.5ab 4.1b 4.1b 5.1a 5.1a <.0001
n = 70 n = 90 n = 102 n = 120 n = 38 n = 30
Salad 5.2a 5.1a 4.8a 4.3b 5.2a 4.9a <.0001
n = 72 n = 98 n = 100 n = 133 n = 27 n = 21
Breadsticks 5.0a 4.6ab 4.4b 4.3b 4.9ab 5.1a 0.0020
 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65 63

3.4. Effects of individual context factors: experimental design. When one is interested in the level
meal and choice of acceptance, then context becomes important.
As with King et al. (2004), not all the context factors
These factors were included in test 3 only. The results evaluated in this study had a significant positive effect on
show a significant (p < 0.0001) and positive effect in the the acceptability of the food items, and not all variables
overall rating of all of the items combined in test 3 (5.0) impacted all foods. Presenting items in the natural envi-
versus tests 1 and 2 (4.1 and 4.0). Similar results were ronment and allowing socialization did not improve the
observed with the individual items (p < 0.0001) as shown overall ratings of the items (test 2 versus test 1). How-
in Table 4, with the exception of lasagna where the ever, evaluating the items under ‘‘natural’’ conditions
scores remained constant across all the tests (4.9, 4.7 (as a patron of the restaurant) had a significant and po-
and 5.0 for tests 1, 2 and 3 respectively). sitive effect on the overall rating of all items except lasa-
gna which remained high but constant.
3.5. Physical environment and social interaction In the case of test 2, the effect of context, specifically
the addition of social interaction and enhanced environ-
The change in physical environment and additional ment, was not evident, suggesting the following possibil-
social interaction between test 1 (laboratory) and test 2 ities: the impact of environment alone and/or social
(restaurant) resulted in no significant change in the over- interaction on the overall acceptability/rating of a prod-
all ratings of the individual (p > 0.34) or combined items uct is not as significant as some of the other context fac-
(p = 0.45). Mean scores dropped between 0.1 and 0.2 tors; the testing approach may have a greater effect on
units (on a 6-point scale) for individual items, except the consumer response than the environment and/or so-
for salad where there was a slight increase (0.1 units), cial interaction; for example, the way the test was exe-
and an overall drop of 0.1 units (4.1–4.0) with all items cuted, including the presentation of the ballot at the
combined. onset of the item evaluation, the way the items were pre-
sented one at a time rather than as part of a meal may all
have contributed to the lower score, overcoming the fact
4. Discussion that they were at the restaurant (natural physical envi-
ronment) consuming the food, or the artificial way so-
A number of studies have demonstrated that the loca- cial facilitation was encouraged in test 2 (consumers
tion where food is tested and/or consumed has an im- were allowed to talk at the table when individuals did
pact on how well that food is liked. Meiselman et al. not know each other). Pliner, Bell, Kinchla, and Hirsch
(2000) demonstrated differences between meals served (2003) demonstrated that social facilitation has a posi-
in institutional settings and restaurant settings, and sug- tive effect on naturally created groups (people that know
gested that laboratory results might fall between these each other) but not on artificially created groups. This
two points. Edwards et al. (2003) confirmed the low rat- would suggest that a meal situation and the ability to
ings in institutional settings and higher ratings in restau- choose items, as well as the way the test is executed such
rant settings, and suggested that consumer expectations as presentation of the ballot, may have a stronger effect
might underlie these differences, based on the report of on the rating of the product than subjects socializing in a
Cardello, Bell, and Kramer (1996) that people expect more natural environment. In the restaurant CLT, the
institutional products to be less well liked. effect of consumer expectation on the menu items did
King et al. (2004) attempted to uncover some of the not overcome all other factors that resulted in low over-
variables underlying the lower scores in laboratories ver- all rating of the products. King et al. (2004) found that
sus restaurants, and found that, generally, consumer socialization may have had a negative effect on the
acceptability of foods increases with the addition of con- acceptability of pizza. In this test, the individual items
text variables. They found that presenting items as part scores dropped slightly except for salad which showed
of a meal and/or giving consumers a choice of items sig- a very slight increase. Because social interaction was cre-
nificantly increased the overall score for several of the ated rather than occurring naturally, participants may
items. have paid more attention to what they were eating in
In the present study, restaurant survey results (test 3) front of others. For example, eating salad might have
were significantly higher than the results of the tradi- been viewed as a healthy lifestyle and therefore created
tional CLT, therefore confirming and extending the re- a positive image in front of strangers, while lasagna
sults of Meiselman et al. (2000) and King et al. (2004). and cannelloni (or pizza) were not viewed necessarily
However, it is important to recognize that the ranking as healthy items, and therefore the scores may not have
of food item scores was not affected by context; in the increased or significantly dropped as in the case of pizza
present study all conditions produced the same ranking (King et al., 2004).
of foods. Thus, if one is only interested in ranking prod- Gender and age effects were evident throughout this
ucts, then context might be less important as a factor in study but, as with context effects, were not consistent
64 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65

throughout all the tests. In general, women appear to previous King et al. (2004) study. This is an important
discriminate among products more than males; how- result because traditional product testing is usually done
ever, differences between genders or among age groups with single food items (not in a meal context), and is
varied by test. Test 1 results showed no significant differ- usually done without the test subject actually selecting
ences between genders or among age group responses the item to be tested. While developing laboratory test
while differences existed in test 2 (restaurant CLT) and protocols using meals rather than items is not difficult,
test 3 (restaurant survey). These results suggest that gen- it is more complex to develop laboratory test protocols
der and age differences may be diminished in laboratory in which the consumer chooses the product to be con-
tests. While this study found differences in gender and sumed and evaluated. Choice is inherently a part of
age for the enhanced settings (Edwards et al., 2003) many natural eating occasions, especially restaurant din-
found the opposite, that there is no difference between ing, which is one reason to use naturalistic testing at cer-
genders in responses to meals in various context settings. tain stages of product development and evaluation.
Therefore, the interaction of context effects with gender Social interaction and environment (location) did not
and age are yet not well understood and additional re- have significant effects on the ratings of the products.
search may be warranted. It is common practice for some restaurant businesses
All of the items tested, with the exception of lasagna, to conduct CLTÕs at the restaurant to create a natural
rated lower in tests 1 and 2 compared with test 3. Lasa- environment/situation for consumers. This study dem-
gna scores remained high for all tests. This high flat onstrated that the facility (restaurant) itself had no mea-
score, may be the result of evaluating a highly familiar sured impact on the acceptance scores and that other
and well liked food product, hence, not showing the factors/approaches may need to be considered to create
contrast found in the other items, particularly cannel- a more natural situation when collecting product data
loni, which scored low in tests 1 and 2 and significantly for restaurants. For those interested in improving the
higher in test 3. A similar pattern was observed with piz- naturalistic setting of tests, it is important to realize that
za in King et al., 2004; scores for pizza were similar and moving an artificial laboratory test to a restaurant or
high between the laboratory CLT and the restaurant. It other natural location might not be sufficient if the test
is possible that well liked and familiar main dishes are protocol remains laboratory-like.
less susceptible to the powerful effects of expectations, This study confirmed the results of King et al. (2004)
which may have more impact on less well-liked foods. that CLT results underestimate consumer judgments of
Well-known and well-liked dishes, such as pizza and products in real-life eating situations. Results of test 1
lasagna (in the US) may have a stronger image, be less versus test 3 in the present test were similar to results
susceptible to being changed by varying eating environ- of test 1 versus test 6 in (King et al., 2004), with meal
ments and the variables included in them. It is also pos- context and food choice having a stronger impact on
sible that products like pizza perform differently in a food acceptability than social interaction or environ-
meal setting. This was suggested by Hedderley and ment. This is an important finding since meal and choice
Meiselman (1995), who observed that pizza accounted elements may be more easily incorporated and better
for a larger portion of overall meal acceptance com- controlled in a CLT approach.
pared to more traditional main dishes. Therefore, we Additional research is still needed to investigate other
cannot conclude that the addition of any context ef- elements present in a CLT to help better predict the
fect(s) in this study result in increased rating scores for acceptability of the product in a real-life situation. Some
all foods. It is recommended that future context studies of those elements include the following: (1) understand-
utilize a broader range of main dishes. Another possibil- ing which foods are more susceptible to which context
ity may be a limitation with the rating scale. Given that effects and (2) whether the context effect is related to
the item is well liked the scores will be high in the initial the initial acceptability of the product (well-liked or
test (test 1) but can not go any higher in test 3 since the familiar foods such as pasta dishes which in some cases
scale is self-limiting. It should also be noted that when stand alone as a complete meal, i.e. lasagna and pizza);
the products were consumed at the restaurant, they were (3) the effect of timing of the ballot presentation (begin-
branded by default. The expected result would have ning of the test versus end of the test); (4) the relative ef-
been a higher score just based on brand disclosure, fect of presenting items as part of a meal versus giving
and since participants frequented the restaurant on a food choices to consumers: are they equally important
regular basis; however this was not the case, suggesting or does one have a greater effect than the other; (5)
once again, that other factors may influence the accept- determining whether the physical environment has an ef-
ability scores. fect once meal and food choice are established; and (6)
Meal and food choice continue to be the strongest fac- investigating whether social facilitation has more or less
tors that influence food perception and acceptability. impact when combined with other factors including
These factors have a strong positive effect on acceptabil- meal duration. Therefore, gaining a better understand-
ity scores for most food items tested during this and the ing of the relative effect and possible ranking of the var-
 S.C. King et al. / Food Quality and Preference 18 (2007) 58–65 65

ious contexts factors investigated thus far (meal, social Edwards, J. S. A., Meiselman, H. L., Edwards, A., & Lesher, L. (2003).
interaction, the environment and food choice) may help The influence of eating location on the acceptability of identically
prepared foods. Food Quality and Preference, 14(8), 647–
determine which may be critical and may need to be 652.
incorporated in consumer tests to better predict ‘‘real Eindhoven, J., & Peryam, D. R. (1959). Measurement of preferences
life’’ responses. for food combinations. Food Technology, 13, 379–382.
Hedderley, E., & Meiselman, H. L. (1995). Modeling meal acceptabil-
ity in a free choice environment. Food Quality and Preference, 6,
15–26.
Acknowledgements King, S., Weber, A., Meiselman, H., & Lv, N. (2004). The effect of
meal situation, social interaction, physical environment and choice
We would like to express our appreciation to Darden on food acceptability. Food Quality and Preference, 15, 645–653.
Restaurants, the parent company of Olive Garden, for Kramer, F. M., Lesher, L. L., & Meiselman, H. L. (2001). Monotony
partnering with McCormick in this study and providing and choice: repeated serving of the same food item to soldiers
under field conditions. Appetite, 36, 239–240.
their expertise as well as the testing location and products Meiselman, H. L. (2002). The effect of context environment on choice
to conduct these studies. We would also like to thank the and acceptance: is context more important than product? Appetite,
chefs at Olive Garden, particularly Chef Ron Grandge- 39, 249 (abstract).
nett, and Chef Kevan Vetter of McCormick & Co., Meiselman, H. L., Hirsh, E. S., & Popper, R. D. (1988). Sensory,
Inc. for assisting in the preparation of the food products. hedonic and situational factors in food acceptance and consump-
tion. In D. M. H. Thomson (Ed.), Food Acceptability (pp. 77–87).
We would also like to thank Roger Thompson at Darden London: Elsevier.
Restaurants, Salli Setta and Mark Gonzalez at Olive Meiselman, H. L., Johnson, J. L., Reeve, W., & Crouch, J. E. (2000).
Garden, as well as Marianne Gillette and Hamed Faridi Demonstrations of the influence of the eating environment on food
at McCormick & Co., Inc. for supporting and encourag- acceptance. Appetite, 35, 231–237.
ing the continuation of this research. Peryam, D. R., & Pilgrim, F. J. (1957). Hedonic scale method of
measuring food preferences. Food Technology, 11, 9–14.
Pliner, P., Bell, R., Kinchla, M., & Hirsch, E. (2003). Time to eat?: the
impact of time facilitation and social facilitation on food intake. In
Pangborn sensory science symposium.
References Rozin, P. (1996). The socio-cultural context of eating and food choice.
In H. L. Meiselman & H. J. H. Macfie (Eds.), Food choice
Beatty, W. W. (1982). Dietary variety stimulates appetite in females acceptance and consumption (pp. 83–104). Glasgow: Blackie Aca-
but not in males. Bulletin of the Psychonomic Society, 19(4), demic and Professional.
212–214. Rozin, P., & Tuorila, H. (1993). Simultaneous and temporal contextual
Bell, R., Meiselman, H. L., Pierson, J. P., & Reeve, W. G. (1994). influences on food acceptance. Food Quality and Preference, 4,
Effects of adding an Italian theme to a restaurant on the perceived 11–20.
ethnicity, acceptability, and selection of foods. Appetite, 22, 11–24. Tuorila, H., Meiselman, H., Bell, R., Cardello, A., & Johnson, W.
Cardello, A. V., Bell, R., & Kramer, F. M. (1996). Attitudes of (1994). The role of sensory and cognitive information in the
consumers toward military and other institutional foods. Food enhancement of certainty and liking for novel and familiar foods.
Quality and Preference, 7(1), 7–20. Appetite, 23(3), 231–246.