Skip to main content
. Author manuscript; available in PMC: 2009 Aug 1.
Published in final edited form as: J Am Diet Assoc. 2008 Aug;108(8):1305–1314. doi: 10.1016/j.jada.2008.05.012

Table 3.

Distribution of Number of Intrusions per Interview by School Meal and Interview Sequence a

School breakfast School lunch
Interview sequence (and number of interviews for which meals reported met criteria to be considered the school meal) Interview sequence (and number of interviews for which meals reported met criteria to be considered the school meal)
1st (n=91) 2nd (n=83) 3rd (n=74) Total (n=248) 1st (n=102) 2nd (n=89) 3rd (n=77) Total (n=268)
Intrusions per interview Number of interviews (% of interviews)
0 16 (18%) 16 (19%) 21 (28%) 53 (21%) 30 (29%) 31 (35%) 34 (44%) 95 (35%)
1 26 (29%) 29 (35%) 24 (33%) 79 (32%) 36 (35%) 27 (30%) 24 (31%) 87 (33%)
2 24 (26%) 16 (19%) 12 (16%) 52 (21%) 20 (20%) 11 (12%) 7 (9%) 38 (14%)
3 15 (16%) 17 (21%) 14 (19%) 46 (19%) 11 (11%) 10 (11%) 9 (12%) 30 (11%)
4 10 (11%) 5 (6%) 2 (3%) 17 (7%) 3 (3%) 7 (8%) 2 (3%) 12 (5%)
5 0 (0%) 0 (0%) 1 (1%) 1 (<1%) 2 (2%) 3 (4%) 1 (1%) 6 (2%)
Interviews with intrusions 75 (82%) 67 (81%) 53 (72%) 195 (79%) 72 (71%) 58 (65%) 43 (56%) 173 (65%)
a

See footnote a of Table 1 for information (about the sample of children, study design, and examples of items classified as intrusions) that is pertinent to Table 3. See footnote b of Table 1 for information (about criteria to be considered a school meal) that is pertinent to Table 3. The breakfast and lunch distributions over all interview sequences differed statistically when they included intrusion-free interviews (P=0.001, chi square), but just missed being statistically different when they excluded intrusion-free interviews (P=0.059, chi square). The proportion of intrusion-free interviews was greater for lunch (95/268=0.35) than breakfast (53/248=0.21) (P<0.001, two-sample test of proportion) but did not vary by interview sequence for breakfast (P>0.206, chi square) or lunch (P>0.122, chi square).