Evaluation on reliability and validity of Chinese version of childhood trauma questionnaire

Xingfu ZHAO; Yalin ZHANG; Longfei LI; Yunfei ZHOU

ABSTRACT

BACKGROUND:

At present, the internationally public recognition of childhood trauma questionnaire is the authorized version of Bernstein,American psychologist, in 1998.

OBJECTIVE:

To set up the scale of Chinese version of childhood trauma questionnaire with 28 items and analyze its reliability and validity.

DESIGN:

Community investigation was designed.

SETTING:

Mental Health Institute, Xiangya Second Hospital, Central South University.

PARTICIPANTS:

Totally 441 students from 8 classes were randomized from a countryside middle school of a city in Henan in October 2004.

METHODS:

A total of 441 students were measured with childhood trauma questionnaire and 93 of them were re-measured 2 months later. Childhood trauma questionnaire included 28 items and divided into 5 subscales,named emotional abuse, physical abuse, sexual abuse, emotional neglect and physical neglect. Five grades were adopted in each item, named 1 score never; 2 scores occasionally; 3 scores sometimes; 4 scores often;and 5 scores always. Every subscale was varied from 5 to 25 scores, and the total results were in the range from 25 to 125 scores. It was to analyze the internal identical property, reliability of re-measuring, average correlation coefficient among items and correlation coefficient between total score and every subscale. And the analysis was carried on validation factors. MAIN OUTCOME

MEASURES:

Homogeneity reliability, re-measuring reliability and validity, absolute fit index, relative fit index and parsimony index of childhood trauma questionnaire were involved.

RESULTS:

Totally 441 pieces of questionnaire were distributed on the spot and 435 pieces of questionnaire with integral and regular answers were collected. Totally 93 pieces of questionnaire were delivered for the second evaluation 2 months later and 93 pieces with regular answers were collected. All of those were used for the evaluation of re-measuring reliability of such measuring table. ① Cronbach α coefficient was 0.64 in Chinese version of childhood trauma questionnaire of and re-measuring reliability was 0.75. Cronbach αcoefficient of every subscale was varied from 0. 16 to 0.65 and re-measuring reliability was in the range from 0.27 to 0.73. The correlation coefficient among items was varied from -0.20 to 0. 44; ②The correlation coefficient between total score and every subscale was varied from 0. 36 to 0.68 and the correlation coefficient among subscales was from-0.01 to 0.39; ③ Indexes of validation factor

analysis:

The load coefficient of physical neglect was varied from 0.09 to 0. 64 and that in 2 items was less than 0.20. Multiple correlation coefficient was varied from-0. 20 to 0. 82, X2/df was 2.48 and root mean square error of appr0ximation(RMSEA) was 0.06. Added indexes IFI (0. 76), CFI (0.75) and TLI (0.72) .Parsimony index PNFI (0.58) and PCFI (0. 67).

CONCLUSION:

Chinese version of childhood trauma questionnaire provides better reliability and validity. According to validation analysis, except physical neglect, every index tallies with psychometric standards. The results of analysis on normative pathways are satisfactory, which explains that the subscales of such measuring model provide good matching property and conception validity.