| Summary: | 碩士 === 東吳大學 === 心理學系 === 106 ===
Food craving has marked relation with obesity. Questionnaires were main tools for assessing food craving. However, questionnaires could not induce participants' appetite nor assessed their immediate responses to food. Moreover, the validity of questionnaires might be influenced by self-awareness and social desirability.
For assessing the automatic prepotent response to food, some researchers developed computerized tests for measuring real implicit attitudes and prepotent responses. However, the dietary behavior has culture bias, these computerized tests are solely applicable to the western society. For establishing an implicit test of dietary bias of Taiwanese, this study employed Implicit Relational Assessment Procedure (IRAP) (Barnes-Holmes, et al., 2006) as the carrier program and select the food stimulus that suitable for Taiwanese.
In the preparatory study, six pictures of high calories foods that can induce the highest level of appetite were selected as the IRAP target stimuli. In contrast, six pictures of low calorie foods that could induce the lowest level of appetite were selected as the other group of IRAP target stimuli. In addition, six vocabularies that had the strongest correction with working up an appetite were selected as label IRAP positive label stimuli. The other six vocabularies that had the strongest correction with inhibit appetite were selected as the other group of IRAP negative label stimuli. After coding the IRAP program, the criterion of maximum response latency and minimum correct rate were chosen via the small-sample pre-test. The study also set up a standard test process for reducing the losing rate.
In the formal study, 144 Taiwanese adults were recruited. The fasting time was set to all participants and the following indices were assessed:
(1) food craving IRAP;
(2) food craving trait questionnaire: trait (FCQ-trait);
(3) Dutch Eating Behavior Questionnaire: external eating (DEBQ-EX);
(4) Dutch Eating Behavior Questionnaire: emotional eating (DEBQ-EMO);
(5) personal information (BMI, age, special dietary restrictions).
The latency data were transformed using the D-algorithm (Barnes-Holmes et al., 2006) for calculating the split-half reliability, criterion-related validity and two kinds of known group validity.
This study used 2 (condition: group) x 4 (D-IRAP category) mixed design to analyze the effects of known group validity.
For testing the food craving trait known group validity, all questionnaire scores were standardized and took 3 times average, then the first quartile (Q1) and the third quartile (Q3) of sample average were chosen as the grouping criterion. The D-IRAP scores of 36 high food craving trait participants and 36 low food craving trait participants were used to data analysis. Two-way ANOVA was used to examine participants ’group across the D-IRAP scores and interaction effect. The simple main effect had been examined subsequently. For testing the BMI known group validity, participants were divided into normal weight and overweight groups in accordance to the BMI cutoff points (BMI>24). The same analyzing methods were used in this part.
Results:
(1) Food craving IRAP presented acceptable split-half reliability;
(2) Low calorie food D-IRAP presented significantly criterion-related validity and two kinds of known group validity;
(3) This study provided a structured testing process to reduce the subject loss rate effectively;
(4) The study provided an accurate stimulus material screening process;
(5) The results of BMI known group validity revealed that overweight group didn't prefer the unappetizing low-calorie food yet accepted it. This dietary tendency was significantly different from the normal weight group, which may offer the new direction for the future
study and test design.
|
|---|
