2 3
4
5
panels on nutritIOn teaching and educatIOn. Journal of NutfltlOn EducatIOn, 1(3):24-32, 1970. Sodowsky, J.D. Inservice nutrition education for elementary teachers. Journal of NutfltlOn EducatIOn, 5:139-42, 1973. Britten, H.C. Nutrition education for elementary education majors. Journal of NutfltlOn EducatIOn, 3:73, 1971. Wakefield, L.M., and A.G. Vaden. Nutrition course for elementary teachers by telephone. Journal of NutfltlOn EducatIOn, 5:190-92, 1973. Callahan, D L. Inservlce teacher
6
7 8
9
workshops. Journal of NutfltlOn EducatIOn, 5:233-36, 1973. Grogan, J. Teacher inservice for nutritIOn educatIOn - An Interdisciplinary approach In the school system. Journal of NutritIOn EducatIOn, 10:119-20, 1978. Head, M.K. A nutrition education program at three levels. Journal of NutritIOn Education, 6:56-59, 1974. Lovett, R. et al. The effect of a nutntlon education program at the second grade level. Journal of NutfltlOn EducatIOn, 2(Supp. 1):80-95, 1970. Parker, E. Big Ideas In nutrition education.
School Foodservlce Journal, 26:50-54, 1972. 10 Shannon, B., P. Bell, E. Marbach, L. O'Connell, K. Graves, and R.F. Nicely, Je. A K-6 nutritIOn curnculum evaluation - InstructIOn and teacher preparation. Journal of NutfltlOn EducatIOn, 13:9-13, 1981. 11 Kuder, G.F., and M.W. Richardson. The theory of estimatIOn of test reliability. Psychometflka, 2:151-60, 1937 12 Cook, C.B., D.A. Elley, and E.C. Kominaha. Too much nutrition education grades K-6? Journal of NutfltlOn EducatIOn, 9:131-35, 1977.
An Inventory for Assessing Food Behaviors of Elementary School Children Alyce M. Fanslow\ Damaris Pease 2 , Shirley C. Gilmore\ and Judy K. Brun 3 Departments of Home Economics Education and 2Chtld Development, Iowa State University, Ames, Iowa 50011; and 3 NatIOnal Datry Council, 6300 North River Road, Rosemont, Illinois 60018 I
This report describes the development, field testing, and summary statiStiCS of a series of Inventories deSigned to assess the food behavIOrs of elementary school students, grades 1 to 6. The content IS that found in elementary school food and nutrItIOn curriculum gUides and focuses upon personal cleanlIness, personal ~afety, and food quality. Students respond to cartoon items by selectmg 1 of 3 responses to indicate their tYPiCal behavior. From results of a national field test With 1,673 students in 75 classrooms, we developed equivalent forms of the inventory. These InventOrIes may serve as a valid and reliable part of a battery of instruments for evaluatIOn of nutrition educatIOn programs. (JNE 14:96-98, 1982)
ABSTRACT
Adequate evaluation of nutrition education programs for elementary school children depends upon the availability of valid and reliable assessment devices that address the goals and objectives of the curriculum. Many available nutrition education evaluation deVices for elementary students assess achievement and focus on concepts such as nutrIents (1, 2). This paper reports the development, field testIng, and summary statistics for eqUivalent forms of an inventory for assessing food behaviors of 96
JOURNAL OF NUTRITION EDUCATION
elementary school students. The items In this instrument assess what students do or intend to do rather than what students know. DEVELOPMENT
We first identified major objectives of the deVice for elementary students by inspectIng 11 food and nutritIon curriculum gUides. From among the common obJectIVe~ we selected one, promotion of healthful habits by using appropriate methods In selectIng, prepanng, and eatIng food, as the
content area for the present assessment deVice. ThiS content area encompasses three major concepts: personal cleanliness, food quality, and personal safety. We then listed speCific food behaVIOrs for each major concept. For example, the behaVIOrs identified for food quality Included: selecting food storage appropriate to the food product, inspecting food quality pnor to using the food, and ensuring food cleanlIness during food preparation. So that the instrument would relate to realistic situations, would be attractive and interesting to the students, would have a standardized format, and would be appropriate for group administration m grades 1 to 6, we selected a cartoon format for the inventory. Each cartoon item illustrates a situation related to selecting, preparing, or eating food and calls for a response by a child to a comment about the situation. The inventory offers three choices for response. These choices include two distractors along with the correct response related to personal cleanliness, food quality, and
VOLUME
14
NUMBER
3
1982
safety. The correct responses selected were those that represented important behaviors based on a scientific principle. The distractors were the two most frequent inappropriate responses obtained by openended interviewing of 120 students. Figure 1 displays cartoon presentatIOn of inventory items. Many of the "incorrect" responses are not necessarily wrong If placed in the context of friendliness, politeness, or other aspects of social interaction. Nonetheless, the "correct" response indicates that the respondent places a higher priority on a response related to safe and healthy food behavior. After content validation of a 50-item preliminary instrument by two nutrition specialists, we pliot tested the mventory with 104 first-grade, 110 third-grade, 100 fourth-grade, and 105 fifth-grade students. The administrator read the cartoons and responses to the students in the first and third grades. The fourth- and fifth-grade students read the cartoons and responses independently. The administrator directed the students to select the response that was most like what they would say in the given situatIOn. Later, under similar conditions we conducted a pilot test for additIOnal or modified cartoon items with 132 additional elementary school students. For each cartoon we calculated the difficulty and discrimination indices (3) by grade. The discriminatIOn mdex is the correlation between a cartoon score and the total inventory score. From the results of the pilot test we found that most cartoons did not differ markedly by grade on difficulty or discrimination. We then factor analyzed the data using the maximum likelihood method, rotating the factors by the Varimax procedure (4) . Results of the factor analYSIS showed that the mventory measured one central theme. That is, while healthful food habits involve behaviors related to personal cleanliness, food quality, and personal safety, the students taking the pilot test did not perceive the specifics within the broader theme. Therefore, we prepared one general inventory for all grade levels. When selectmg Items for the field test version of the mstrument, we gave preference to items with difficulty mdlces between 30070 and 70070 and discrimmatlon mdices greater than 0.2, while retaming representatIOn of the three content areas. In all, we Identified 54 cartoons for Illclusion in the field test version of the inventory.
VOLUME
14
NUMBER
3
1982
JI
b tf
Where should I put the leftover mac droni and ~ cheese?
1 1~?/~'~.@A. .QI -.
~~-'
ti Q
In the cupboard .
B.
On the counter.
C.
In the refrLgerator.
) This box is chewed
2./
_\-:)-+c---------1.---+:7 (3
A.
Put it in another box.
B.
Try to fix it.
C.
Throw it away .
I'll eat it anyway I should get another bun. C
Figure 1
I'll brush it off.
Examples of cartoon Items from th e food behaVIOr Inventory
FIELD TEST
We conducted a field test on the mstrument in 75 classrooms With 1,673 students in grades I to 6. To obtain a diverSity of backgrounds among the students, we Illeluded III the field test 12 schools III 10 different states representmg low, middle, and high socIOeconomic levels. To determme socioeconomic level, we considered the percentage of students receiving federally subsidized lunches, the neighborhood
housing density, and the occupatIOnal level of parents. Boys represented 49070 of the sample and girls 51070; 77070 were white, 21070 were black, and 2UJo were other minonties. In the field test, children at all grade levels completed 54 cartoon items. However, so that younger students would not tire before completing all the cartoons, students III grade I completed approxImately one-third of the cartoons at one time; students III grades 2 and 3 completed
JOURNAL OF NUTRITION EDUCA TION
97
Table 1
I nventory statistics by grade and form Grade Level
Slal/slies
RehabJ!lly Internal consIstency EquIvalence Mean S~ore Standard devIatIon Average dIfficulty Number of cartoons
2
3
(n = 256)u
(n = 263)
(n = 273)
AI!
A
A
B
B
B
5
4 (n = 3/1)
(n = 292)
A
A
B
B
6 (n = 278)
A
B
075
076
0.75
0.76
0.78
080
082 081
0.85
085
0.87
0.73
0.73
0.78
0.78
080 080
084 0.84
0.85
085
0.89 0.89
088
12.2
124
14.7
149
175
17.5
17 .2
169
186
183
18 .9
18.6
36
3.6
36
36
42
4.2
4.7
46
52
53
5.5
58
68070 69%
73% 75%
73070 73%
72 % 71 %
18
20
24
24
18
20
24
24
72% 70% 26
26
73% 72% 26
26
.In = number of , tudent, partlclpatmg In the field testmg at each grade level bTest form A or B
approxImately one-half; and students in grades 4, 5, and 6 completed all 54 cartoons in one session. Each test requested the sex and grade level of the student and contaIned a code number to ensure the anonymIty of schools and students. The Iowa State UnIversity CommIttee on the Use of Human Subjects In Research reviewed and approved the Inventory and the plan for the field test. We computed dIfficulty and dlscnmInatlOn IndIces for each cartoon by grade level. One of the cartoons had a negatIve discriminatIOn Index, and another one was too dIfficult. Therefore, these two cartoons were deleted . All other cartoons had acceptable dIfficulty and dlscnmmation IndIces. From the results of the field test, we created equivalent forms of the Inventory for dIfferent age groups. We arranged all cartoons accordIng to difficulty mdlces and selected two that assessed the same concept area and were of approxImately equal dIfficulty for parallel Items on the eqUIvalent Inventory forms, Form A or Form B. After studying the summary statIStICS for equIvalence, we deCIded to reassIgn cartoons to the two forms to obtain simIlar summary statIstIcs as well as acceptable levels of the reliablltty coefficients. The equivalent forms vary in number of cartoons by grade. Grade 1 has 18 cartoons; grade 2, 20 cartoons; grades 3-4, 24 cartoons; and grades 5-6, 26 cartoons. At each grade level, approxImately one-half of the cartoons address food quality and approximately one-fourth of the cartoons address each of the other two content areas : personal cleanliness and personal safety.
98
JOURNAL OF NUTRITION EDUCATION
THE FINAL VERSION Table 1 presents the summary statistics for the equivalent forms of the inventories at each grade level. The coefficients of Internal consistency are computed by the Kuder-RIchardson Formula 20 (KR-20). The coefficient of equivalence IS the correlation coefficient between the two mventory forms at each grade level. These inventories (see Note) should be suitable for use in a variety of situations because they possess the characteristIcs of content validity, objectivity, and reliability. A reliabilIty coefficient of 0.75 or above (5) is considered acceptable for group measurement. This numeric value of the coefficient is also acceptable for individual measurement when a score is used in conjunction with other measures of student performance. For thIS mventory, the coeffiCients of internal consistency were 0.75 or above for all grade levels, and coefficients of equivalence were all above 0.75 except for grade 1, where the mstrument contained the fewest cartoons. Thus, WIth the possIble exceptIOn of the grade 1 instrument, these mventones can be used for the two purposes described. The statistical data for the lllventories provide a basis for interpreting individual or group performance relative to other groups who have responded. The inventory difficulty is somewhat easier than the optImum inventory difficulty (5) of 670/0the point midway between the chance probabilIty of success (i.e., 33%) and 100%. However, the inventories did discriminate between high and low scorers. The mean score and standard deVIation in Table 1 can be used to provide a reference point when using the inventories to evaluate the food behaviors of students.
Since these statistics derive from field test scores from dIverse settings, they should be appropriate for comparison with most elementary classrooms in the United States. The availability of equivalent forms of the inventories allows a flexibility in developing evaluation design. We believe that these inventories will add a useful new dimension to evaluation of nutrition edu0 cation in elementary schools.
ACKNOWLEDGMENTS
This research was funded In part by the National DaIry CouncIl. ThiS paper IS Journal Paper No. 237 of the Home EconomIcs Research Institute, College of Home EconomIcs, Iowa State UniveTSlty, Project No. 178. NOTE
Flxmg and Eatmg My Food, food behavior inventones for grades 1 to 6, WIll soon be available from the NatIOnal Dairy CounCIl, 6300 North RIver Road, Rosemont, IllInois 60018 LITERATURE CITED
1 Cooper, K . A., and C. E. Go. AnalYSIS of
nutnuon education curnculum gUIdes at the
K-12 level. Journal oj NutritIOn EducatIOn, 8:62-66, 1976. 2 Levy, S. R., B. K. Iverson, and H. J. WaJ-
berg. NutntlOn education research: An interdISCIplinary evaluatIOn and revIew. Health
EducatIOn Quarterly, 7:107-26, 1980. 3 American PsychologIcal ASSOCIatIOn. Standards jor educatIOnal and psychological tests. Washmgton, D.C.: Amencan PsychologIcal ASSOCIatIOn, 1974, pp. 25-55 . 4 Harman, H H. Modern Factor AnalYSIS. ChIcago. Umverslty of ChIcago Press, 1962, pp. 301-308.
5 Tmkleman, S. N. Planning the objecttve test. In EducatIOnal Measurement, 2d ed., R. L.
Thorndike, ed WashIngton, D.C.. American CounCIl on EducatIOn, 1971, pp. 62, 71.
VOLUME
14
NUMBER
3
1982