Food Quality and Preference 16 (2005) 131–138

www.elsevier.com/locate/foodqual

Consumer testing with children on food combinationsfor school lunch

Ella Pagliarini a,*, Nicola Gabbiadini a, Sabrina Ratti b

a DISTAM, Sezione Tecnologie Alimentari, Universit�a degli Studi di Milano, via Celoria 2, 20133 Milano, Italyb Studio Associato Gemini, Via Chianciano 9, 20161 Milano, Italy

Accepted 3 March 2004

Available online 13 April 2004

Abstract

Awareness of how important food and food culture are to the quality of children’s life is a presupposition that promotes research

in the sector of school lunch. The aim of this work was to evaluate liking for meals supplied to primary school refectories of the

Municipality of Milan. It appears particularly interesting to investigate consumption of meals in Italy in view of both food com-

binations and the nutritional aspect of the Mediterranean diet. In addition, meals are typically subdivided into four distinct

components in Italy. All the children (120) involved, aged between 7 and 10, supplied liking scores as shown in a previous work

[Italian J. Food Sci. 2 (2003) 215] for a set of first courses, second courses, vegetables and fruit/dessert.

From results it can be seen that preferences for most dishes of younger children (7 years old) differed from those of the older ones

(10 years old). Also, unlike 8-, 9- and 10-year-old children, 7-year-old children supplied steadily higher acceptability scores than the

intermediate score on the evaluation scale used. Based on our data, it was shown that children become increasingly aware of their

preferences and critical in their choices with growing age.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Children; School lunch; Consumer test; Acceptability

1. Introduction

Awareness of how important food and food culture

are to the quality of life of people, especially children,

not only in terms of health but also from a social, cul-

tural point of view, is a presupposition that promotes

research in the sector of school lunch in Italy (Pagliarini,Ratti, Balzaretti, & Dragoni, 2003).

Beyond the intrinsic quality of food preparations,

liking for school lunch by children is affected by several

various factors, which are mainly connected to envi-

ronmental conditions of use, absence of a food educa-

tion program, influence on assessment developed both

among children and with a leader, influence deriving

from a comparison with home lunch, and also relatingto how familiar children are with a specific food product

(Birch, Birch, Marlin, & Kramer, 1982; Leon, Cour-

*Corresponding author. Tel.: +39-25031-6620; fax: +39-25031-

6632.

E-mail address: ella.pagliarini@unimi.it (E. Pagliarini).

0950-3293/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.foodqual.2004.03.001

onne, Marcuz, & Koster, 1999; Schraidt, 1991; Spaeth,

Chambers, & Schwenke, 1992).

Parents, teachers, public officials, and restaurateurs

are often responsible for making available the food

items consumed by children; in many situations the

preferences of the child are considered and followed, in

many others they are not. In particular, economic andnutritional factors determine the options open to chil-

dren in the public sector of the school system (Ahlstrom,

Baird, & Jonsson, 1990; Noble, Corney, Eves, Kipps, &

Lumbers, 2000).

Having lunch at school has an important educational

function because the diet implies a number of hidden

significances, namely a physiological significance to

learn to feed properly (Bell�u & Cucco, 1997; Bell�u et al.,1995), a cultural significance to know different varieties

and origins of foods (Baxter, Thompson, & Davis,

2000), and a psychological significance to understand

why a specific food product may arouse emotions

(Kimmel, Sigman-Grant, & Guinard, 1994). Impor-

tantly, healthy eating habits established in childhood

can reduce the risk of developing diet-related diseases in

132 E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138

adulthood, by which time the re-education of one’s

dietary habits may be difficult to maintain. Additionally,

it has been suggested that children are more receptive to

dietary changes than are adults, so there may be awindow of opportunity which unhealthy dietary prac-

tices can be corrected during childhood (Baxter, Sch-

roder, & Bower, 2000).

Therefore, it appears particularly interesting to

investigate consumption of school meals by children in

Italy because the Mediterranean diet is important as a

result of its food combinations, nutritional aspects

and subdivision into the following four compo-nents: first course, second course, vegetables and fruit/

dessert.

On the other hand, very few studies concerning con-

sumption of meals by primary school children have been

published in Italy (Caporale, Cantore, & Monteleone,

2003; Monteleone, Caporale, Cantore, & Carlucci, 2003;

Pagliarini et al., 2003).

The aim of this work was to evaluate liking for mealssupplied to school refectories of the Municipality of

Milan in agreement with Pagliarini et al. (2003). Eight

classes from the same school were chosen for a total of

120 children. Children involved provided an accept-

ability score for a set of first courses, second courses,

vegetables and fruit/desserts.

Table 1

List of the formulations tested

Formulations First course Second course

A Penne with pesto Cheese

B Risotto with pumpkin Roasted turkey

C Pasta with tomato sauce Roasted pork loin

D Barley soup Chicken breast

E Vegetable pur�ee with rice Vegetable omelette

F Fusilli with oil and parsley Turkey with tomato

oregano sauce

H Risotto with tomato Roasted pork with a

sauce

I Ravioli with butter and sage Cooked ham

J Pasta and bean soup Roasted turkey

K Pasta with tomato and oregano

sauce

Vegetable omelette

L Vegetable soup Dried salted beef

M Risotto with saffron Fish fingers

N Risotto with tomato Turkey with tomato

oregano sauce

O Pasta with tomato sauce Small Mozzarella che

balls

P Risotto with pumpkin Hake with tomato sa

Q Creamed vegetable soup and

croutons

Cooked ham

R Pasta with tomato and Ricotta

cheese

Roasted turkey

T Pasta with tomato sauce Roasted pork loin

U Pasta with legumes Roasted pork loin

X Risotto with beet tops Roasted turkey

Y Vegetable soup Roasted pork loin

Z Pizzoccheri with potatoes –

Moreover, we were interested in investigating the role

played by different food components in determining

acceptance of different meals and evaluating how the

factor age may affect children’s preferences.

2. Materials and methods

2.1. Sample selection

Formulations (i.e. first course, second course, vege-

tables and fruit/dessert) reflecting the meals evaluated

were reported in Table 1.

2.2. Subjects

For panel formation eight classes from the same

school were selected as follows: two second-, two third-,

two fourth- and two fifth-grade classes consisting of 15

pupils each for a total of 120 children.

All of the children, aged between 7 and 10, invariably

had lunch at 12:30–13:00.

2.3. Methodology

A 7-point facial hedonic scale from super good (7) to

super bad (1) was chosen to identify liking for the four

Vegetables Fruit/dessert

Boiled potatoes (with parsley) Apple

Buttered French beans Pear

Mashed potatoes Orange

Boiled zucchini Pear

Red salad and fennels Apple

and Mashed potatoes Pear

pple Grated carrots Orange

Boiled zucchini Apple

Boiled potatoes (with parsley) Apple

Tomatoes Apple

Green salad and carrots Peach juice

Boiled potatoes (with parsley) Pear

and Mashed potatoes Pear

ese Tomatoes Apple

uce Boiled potatoes (with parsley) Banana

Tomatoes Apple

Boiled potatoes (with parsley) Pear

Mashed potatoes Apple

Tomatoes Peach juice

Mashed potatoes Apple

Boiled potatoes (with parsley) Apple

Buttered French beans Chocolate pudding

E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138 133

meal components according to Pagliarini et al. (2003).

Tests started in April and ended in June 2002 accounting

for a total of 22 sessions and 4600 filled in cards.

2.4. Data analysis

Sensory data were submitted to one-way Analysis of

Variance (ANOVA), by applying the Scheff�e test

(p < 0:05) to verify whether there were significant dif-ferences between formulations. The significance of ef-

fects was tested with F test by Statgraphics� Plus

(version 5) software.

3. Results and discussion

In our previous study it was demonstrated that chil-dren of the same age from different grades showed

homogeneous preferences.

In the processing below, the variable grade member-

ship was not taken into account, and the variable age

was only considered in the following way: 7 years (sec-

ond-grade classes), 8 years (third-grade classes), 9 years

(fourth-grade classes) and 10 years of age (fifth-grade

classes). Two classes were taken into account in order toexamine at least 30 children per age category. It was first

investigated whether our untrained young assessors may

be reliable in assigning scores to meals. Since the

experimental design did not include replicates of menus

served, replicates of repeated dishes (risotto with

pumpkin, pasta with tomato sauce, risotto with tomato,

vegetable soup, roasted pork loin, roasted turkey,

cooked ham, mashed potatoes and boiled zucchini),except for fresh raw vegetables and fruit/dessert because

of their intrinsic variability, on different days were

studied in order to check repeatability of assessors. Data

were subjected to both t test (p < 0:05), when two rep-

licates were performed, and one-way Analysis of Vari-

ance (ANOVA), by applying the Scheff�e test (p < 0:05)to verify whether there were significant differences be-

tween dishes.Most repeated dishes, except for risotto with tomato

(F8 years ¼ 10:85���; F10 years ¼ 6:84��), roasted pork loin

(F9 years ¼ 27:90���); cooked ham (F9 years ¼ 15:85���;F10 years ¼ 24:35���) and mashed potatoes (F8 years ¼3:61��; F9 years ¼ 5:24���) showed no significant differ-

ences. Therefore, these untrained young consumers may

be considered reliable on the whole.

Results from 7-to-10-year-old children were thuscompared by one-way analysis of variance to verify

whether there were significant differences according to

the age.

With reference to first courses, in order to facilitate a

comparison between formulations B and P (corre-

sponding to risotto with pumpkin), C, O and T (corre-

sponding to pasta with tomato sauce), H and N

(corresponding to risotto with tomato), L and Y (cor-

responding to vegetable soup), a mean was first calcu-

lated, and all of the formulations were then evaluated

together. As a result, the total of samples tested corre-sponded to 17 first courses.

In Figs. 1–4, a line was drawn at point 4, which

marked the ‘‘maybe good’’ or ‘‘maybe bad’’ score (mean

data point on the evaluation scale).

Fig. 1 shows results from a comparison between first

courses.

Generally speaking, the most preferred first courses

appeared to be: risotto with pumpkin (samples B and P),fusilli with oil and parsley (sample F), risotto with to-

mato (samples H and N), ravioli with butter and sage

(sample I), pasta with tomato and oregano sauce

(sample K), risotto with saffron (sample M), creamed

vegetable soup with croutons (sample Q) and risotto

with beet tops (sample X). The most disliked first

courses appeared to be: barley soup (sample D) and

pasta and bean soup (sample J). It can also be observedthat, unlike older children, 7-year-old children invari-

ably provided a higher score than 4 for all first courses.

It may be suggested that older children may be more

self-confident in their preferences and, hence, make

critical choices. In addition, the behaviour of younger

children (7 years old) differed from that of the older ones

(10 years old) for all first courses that showed a signif-

icant difference (p < 0:05).In order to obtain a detailed investigation, children’s

preferences were further compared to verify whether

there were significant differences according to the age by

classifying first courses into two groups as follows:

soups (samples D, E, J, L, Q and Y) and solids (samples

A, B, C, F, H, I, K, M, N, O, P, R, T, U, X and Z).

Finally, samples were further subdivided into pasta-

based (samples A, C, F, K, O, R, T and U) and rice-based dishes (samples B, H, M, N, P and X), soups

(samples D, E, J, L, Q and Y) and the remaining dishes

(samples I and Z).

Results from comparisons were reported in Table 2.

Both comparisons invariably showed significant dif-

ferences in liking for first courses, except for dishes I and

Z. During data processing the behaviour of 7-year-old

children was again very different from and their likingfor all samples higher than that of 10-year-old children.

As mentioned previously, this behaviour may be ex-

plained by the fact that children become increasingly

selective with growing age.

With reference to second courses, in order to facilitate

a comparison between formulations B, R, X and J

(corresponding to roasted turkey), C, T, U and Y

(corresponding to roasted pork loin), E and K (corre-sponding to vegetable omelette), F and N (correspond-

ing to turkey with tomato and oregano sauce), I and Q

(corresponding to cooked ham), a mean was first cal-

culated, and all of the formulations were then evaluated

1

2

3

4

5

6

7

A BP COT D E F HN I J K LY M Q R U X Z

FIRST COURSES

PR

EF

ER

EN

CE

SC

OR

ES

7 year-old 8 year-old

9 year-old 10 year-old

a

a

ab

b

a abab

b

bb

a

a

a

ab

b

b

a

a

ab

b

b

bb

a

b

b

ab

a

a

ab

abb bc

c

ab

a

b

b

a

a

Fig. 1. Histograms for liking for first courses with respect to different classes. Bars marked with different letters are significantly different (p < 0:05).

1

2

3

4

5

6

7

A BRXJ CTUY D EK FN H IQ L M O P

SECOND COURSES

PR

EF

ER

EN

CE

SC

OR

ES

7 year-old 8 year-old

9 year-old 10 year-old

a

ab

b aba

abb

ab

a

abb

aba

ab

b

ab

Fig. 2. Histograms for liking for second courses with respect to different classes. Bars marked with different letters are significantly different

(p < 0:05).

134 E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138

together. As a result, the total of samples tested corre-

sponded to 12 second courses, as shown in Fig. 2.

From data observation, it can be seen that there were

significant differences in the following second courses

according to the age: cheese (sample A), roasted turkey

(samples B, R, X and J), roasted pork loin (samples C,

T, U and Y) and vegetable omelette (samples E and K).

Therefore, preference for second courses resulted to be

more homogeneous among differently-aged children.

Generally speaking, the most preferred second courses

appeared to be: roasted pork loin (samples C, T, U and

Y), roasted pork with apple sauce (sample H), cooked

1

2

3

4

5

6

7

AMPRYJ BZ CFNTX DI E H L OQUK

VEGETABLES

PR

EF

ER

EN

CE

SC

OR

ES

7 year-old 8 year-old

9 year-old 10 year-old

a

ab

ab b

a

ab

bb

aa

b

a

a

ab b

a

b

a

ab

b

b

bb

ab

Fig. 3. Histograms for liking for vegetables with respect to different classes. Bars marked with different letters are significantly different (p < 0:05).

1

2

3

4

5

6

7

AEIOQTXYKJ DFMNR LU CH ZP

FRUIT/DESSERT

PR

EF

ER

EN

CE

SC

OR

ES

7 year-old 8 year-old

9 year-old 10 year-old

bbab

a

aa a

b

Fig. 4. Histograms for liking for fruit/dessert with respect to different classes. Bars marked with different letters are significantly different (p < 0:05).

E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138 135

ham (samples I and Q), dried salted beef (sample L), fish

fingers (sample M), small Mozzarella cheese balls

(sample O). The most disliked second course appeared

to be cheese (sample A), although 7- and 8-year-old

children assigned to this dish a higher score than 4.

In order to obtain a detailed investigation, children’s

preferences were further compared to verify whether

there were significant differences according to the age by

classifying second courses into two groups as follows:

cold (samples A, I, L, O and Q) and hot dishes (samples

B, C, D, E, F, H, K, M, N, P, R, T, U, X, Y and J). In

addition, samples were further subdivided into meat

(samples B, C, F, H, N, R, T, U, X, Y and J), fish

(samples M and P), cheese (samples A and O), eggs

Table 2

Means for acceptability for different subgroups of first courses

Significance of subgroup for first course 7 years 8 years 9 years 10 years

Soups (F ¼ 36:86���) 5.39a 4.79b 4.04c 3.36d

Solids (F ¼ 45:60���) 5.50a 4.90b 4.39c 4.12c

Pasta (F ¼ 32:84���) 5.33a 4.83b 4.02c 3.67c

Rice (F ¼ 18:60���) 5.82a 4.89b 4.72b 4.62b

Soups (F ¼ 36:86���) 5.39a 4.79b 4.04c 3.36d

Other (F ¼ 2:75 n.s.) 5.18 5.22 4.70 4.20

Within row values marked with different letters are significantly different (n.s. and ��� denotes not significant and significant at 0.001, respectively).

136 E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138

(samples E and K) and charcuterie (samples I, L and Q).Finally, meat-based samples were further subdivided

into turkey (samples B, F, N, R, X and J), chicken

(samples D) and pork meat (samples C, H, I, Q, T, U

and Y).

Results from comparisons were reported in Table 3.

As it can be observed, there were significant differ-

ences in the following second course subgroups

according to the age: hot dishes, meat, eggs, turkey andpork meat. In this case, too, except for the subgroups

meat and turkey, the behaviour of 7-year-old children

differed from that of 10-year-old children.

With reference to vegetables, in order to facilitate a

comparison between formulations A, M, P, R, Y and J

(corresponding to boiled potatoes), B and Z (corre-

sponding to buttered French beans), C, F, N, T and X

(corresponding to mashed potatoes), D and I (corre-sponding to boiled zucchini), O, Q, U and K (corre-

sponding to tomatoes), a mean was first calculated, and

all of the formulations were then evaluated together. As

a result, the total of samples tested corresponded to 8

vegetables, as shown in Fig. 3.

From data observation, it appeared that six out of

eight vegetables showed significant differences according

to the age. Except for mashed potatoes (sample C, F, N,T and X), the behaviour of 7-year-old children, who

invariably provided a higher score than 4 for all vege-

tables, always differed from that of older children.

Table 3

Means for acceptability for different subgroups of second courses

Significance of subgroup for second course 7 years

Cold dishes (F ¼ 0:72 n.s.) 5.40

Hot dishes (F ¼ 7:73���) 5.42a

Meat (F ¼ 6:52���) 5.47a

Fish (F ¼ 0:17 n.s.) 5.37

Cheese (F ¼ 0:17 n.s.) 5.13

Eggs (F ¼ 3:40�) 5.22a

Charcuterie (F ¼ 1:76 n.s.) 5.57

Turkey (F ¼ 3:89��) 5.20a

Chicken (F ¼ 1:92 n.s.) 5.70

Pork (F ¼ 4:00��) 5.77a

Within row values marked with different letters are significantly different (n.

respectively).

Generally speaking, the most preferred vegetables ap-peared to be: green salad and carrots (sample L),

tomatoes (samples O, Q, U and K), mashed potatoes

(samples C, F, N, T and X). The most disliked vegeta-

bles appeared to be: boiled zucchini (samples D and I)

and red salad and fennels (sample E).

In order to obtain a detailed investigation, children’s

preferences were further compared to verify whether

there were significant differences according to the age byclassifying vegetables into two groups as follows: raw

(samples E, H, L, O, Q, U and K) and cooked vegetables

(samples A, B, C, D, F, I, M, N, P, R, T, X, Y, J and Z).

In addition, samples were further subdivided according

to the vegetable type into differently prepared potatoes

(samples A, C, F, M, N, P, R, T, X, Y and J), French

beans (samples B and Z), zucchini (samples D and I),

tomatoes (samples O, Q, U and K), salad (samples Eand L) and carrots (sample H).

Results from comparisons were reported in Table 4.

As it can be seen, there were significant differences in

all the above-mentioned vegetable subgroups, except for

salad (samples E and L). Also, the behaviour of 7-year-

old children invariably differed form that of 10-year-old

children.

With reference to fruit/dessert, too, in order to facil-itate a comparison between formulations A, E, I, O, Q,

T, X, Y, K and J (corresponding to apple), B, D, F, M,

N and R (corresponding to pear), L and U (corre-

8 years 9 years 10 years

5.29 5.11 5.17

5.22ab 4.87b 4.99b

5.32a 4.87b 5.11ab

5.22 5.17 5.15

5.28 4.91 4.38

4.68ab 4.58ab 3.95b

5.29 5.21 5.70

5.20a 4.55b 5.06ab

4.87 4.93 4.77

5.47ab 5.26b 3.32b

s., �, ��, ��� denotes not significant, significant at 0.05, 0.01, and 0.001,

Table 4

Means for acceptability for different subgroups of vegetables

Significance of subgroup for vegetables 7 years 8 years 9 years 10 years

Raw vegetables (F ¼ 9:64���) 5.42a 4.89ab 4.59b 4.48b

Cooked vegetables (F ¼ 18:91���) 5.12a 4.74a 4.18b 4.26b

Potatoes (F ¼ 9:72���) 5.38a 5.13ab 4.54c 4.89bc

French beans (F ¼ 6:69���) 4.93a 4.17ab 3.36b 3.48b

Zucchini (F ¼ 7:09���) 3.97a 2.87ab 2.80b 2.33b

Tomatoes (F ¼ 4:93��) 5.55a 5.24ab 4.68b 4.78b

Salad (F ¼ 1:54 n.s.) 5.08 4.52 4.43 4.32

Carrots (F ¼ 6:58���) 5.80a 4.40b 4.50ab 3.77b

Within row values marked with different letters are significantly different (n.s., ��, ��� denotes not significant, significant at 0.01 and 0.001,

respectively).

E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138 137

sponding to peach juice), C and H (corresponding to

orange), a mean was first calculated, and all of the for-mulations were then evaluated together. As a result, the

total of samples tested corresponded to 6 fruit/desserts.

From data in Fig. 4 it can be observed that a high

score for all fruit/dessert samples was assigned, and

homogeneous preferences were obtained regardless of

the age; significant differences were only found in apple

and pear.

Fruit/dessert samples were not further subdividedbecause of the restricted number of subgroups.

4. Conclusions

From data analysis it can be firstly noticed that rep-

licates of most dishes on different days did not show

significant differences in acceptability. These resultsdemonstrated that evaluation cards were effective, and

assessors were reliable, despite their young age.

Secondly, from average acceptability scores it ap-

peared that the most preferred first courses included

solids, especially rice-based dishes; among second

courses the highest acceptability scores were obtained by

charcuterie and meat-based dishes; finally, cooked veg-

etables were preferred to raw vegetables.With reference to the difference in age, it was ob-

served that children’s preferences changed during pri-

mary school years; in particular, younger children (7

years old) provided higher acceptability scores for most

formulations than older children (10 years old), who

assigned lower, more varied scores. It may therefore be

assumed that children become increasingly more critical

in their food choices with growing age.Further research will have to be carried out to go into

this issue in detail, taking into account that school

refectories should develop into children-oriented res-

taurants. Using proper cooking methods and suitable

ingredient combinations should, therefore, result in

more pleasant recipes and more balanced dishes for

children eating at refectories of Italian schools.

Acknowledgements

We are grateful to Milano Ristorazione SpA staff for

their collaboration and technical support.

References

Ahlstrom, R., Baird, J. C., & Jonsson, I. (1990). School children’s

preferences for food combinations. Food Quality and Preference, 2,

155–165.

Baxter, I. A., Schroder, M. J. A., & Bower, J. A. (2000). Children’s

perceptions of and preferences for vegetables in the west of

Scotland: the role of demographic factors. Journal of Sensory

Studies, 15, 361–381.

Baxter, S. D., Thompson, W. O., & Davis, H. C. (2000). Fourth-grade

children’s observed consumption of, and preference for, school

lunch foods. Nutrition Research, 20(3), 439–443.

Bell�u, R., & Cucco, I. (1997). Determination of intra-/inter-individual

variability and its effect on the number of days required to assess

the usual intake of a school children population. Nutrition

Research, 17(11/12), 1655–1662.

Bell�u, R., Ortisi, M. T., Riva, E., Banderali, G., Cucco, I., &

Giovannini, M. (1995). Validity assessment of a food frequency

questionnaire for school-age children in Northern Italy. Nutrition

Research, 15(8), 1121–1128.

Birch, L. L., Birch, D., Marlin, D. W., & Kramer, L. (1982). Effects of

instrumental consumption on children’s food preference. Appetite:

Journal for Intake Research, 3, 125–134.

Caporale, G., Cantore, C., & Monteleone, E. (2003). Ottimizzazione e

validazione di un metodo per lo studio delle preferenze alimentari di

soggetti in et�a prescolare. In Proceedings of the Ruolo dell’analisi

sensoriale per la valorizzazione delle produzioni alimentari italiane

symposium, no. C13, Roma, Italy, 13–14 November.

Kimmel, S. A., Sigman-Grant, M., & Guinard, J. X. (1994). Sensory

testing with young children. Food Technology, 48(3), 92–99.

Leon, F., Couronne, T., Marcuz, M. C., & Koster, E. P. (1999).

Measuring food liking in children: a comparison of non verbal

methods. Food Quality and Preference, 10, 93–100.

Monteleone, E., Caporale, G., Cantore, C., & Carlucci, A. (2003).

Optimization methods for the study of food preferences in children.

In Poster at 5th Pangborn sensory science symposium, A Sensory

Revolution, no. P146, Boston, USA, 20–24 July.

Noble, C., Corney, M., Eves, A., Kipps, M., & Lumbers, M. (2000).

Food choice and school meals: primary schoolchildren’s percep-

tions of the healthiness of foods and the nutritional implications of

food choices. Hospitality Management, 19, 413–432.

Pagliarini, E., Ratti, S., Balzaretti, C., & Dragoni, I. (2003). Evalu-

ation of a hedonic scaling method for measuring the acceptability

138 E. Pagliarini et al. / Food Quality and Preference 16 (2005) 131–138

of school lunches by children. Italian Journal of Food Science,

2(15), 215–224.

Schraidt, M. F. (1991). Testing with children: getting reliable infor-

mation from kids. In ASTM Standardization News, March, pp.

42–45.

Spaeth, E. E., Chambers, E., IV, & Schwenke, R. A. (1992). A

comparison of acceptability scaling methods for use with children.

In Wu/Gelinas (Ed.), Product testing with consumers for research

guidance: special consumer groups (Vol. 2, pp. 65–77). Philadelphia:

ASTM Publication.