วารสารสมาคมจิตแพทย์แห่งประเทศไทย :: JOURNAL OF THE PSYCHIATRIC ASSOCIATION OF THAILAND

วารสารสมาคมจิตแพทย์แห่งประเทศไทย
Journal of the Psychiatrist Association of Thailand
ISSN: 0125-6985
บรรณาธิการ มาโนช หล่อตระกูล
Editor: Manote Lotrakul, M.D.

การสร้างเกณฑ์ปกติสำหรับวัดสุขภาพจิตในคนไทย

สุชีรา ภัทรายุตวรรตน์ ปร.ด.*
เธียรชัย งามทิพย์วัฒนา พ.บ.*
กนกรัตน์ สุขะตุงคะ วท.ม.*

*ภาควิชาจิตเวชศาสตร์ คณะแพทยศาสตร์ศิริราชพยาบาล มหาวิทยาลัยมหิดล กรุงเทพมหานคร 10700

การสร้างเกณฑ์ปกติสำหรับแบบวัดสุขภาพจิตในคนไทย เพื่อให้ได้เกณฑ์ปกติสำหรับเป็นเกณฑ์ในการวิเคราะห์และการตีความหมายผลการทดสอบ ใช้กลุ่มตัวอย่างสำรวจจำนวน 700 ราย โดยกลุ่มตัวอย่างแยกเป็น กลุ่มคนปกติ กลุ่มผู้ป่วยจิตเวชนอก และผู้ป่วยจิตเวชใน เครื่องมือวัดสุขภาพจิตที่พัฒนาขึ้นสร้างตามเกณฑ์การวินิจฉัยของ DSM-IV ผลการศึกษาพบว่า แบบสอบถามที่ทำการพัฒนาขึ้นมีค่าอำนาจจำแนกระหว่างคนปกติและกลุ่มผู้มีปัญหาทางจิตเวชอย่างมีนัยสำคัญทางสถิติที่ระดับ .001 และดำเนินการสร้างเกณฑ์ปกติด้วยคะแนนมาตรฐานที

วารสารสมาคมจิตแพทย์แห่งประเทศไทย 2542; 44(4): 285-297.

คำสำคัญ แบบวัดสุขภาพจิตสำหรับคนไทย เกณฑ์ปกติ

Sucheera Phattharayuttawat, Ph.D.*
Thienchai Ngamthipwattana, M.D.*
Kanokrat Sukhatungkha, M.Sc.*

*Department of Psychiatry, Faculty of Medicine Siriraj Hospital, Bangkok 10700, Thailand.

The purpose of this study was to develop the Norm Profile for the Thai Mental Health Questionnaire. The survey samples were 700 samples consisted of normals, psychiatric outpatients, and psychiatric inpatients. Data collected from the Thai Mental health Questionnaire was developed from criteria based on the DSM-IV. The results of this study showed that the final test composed of 70 items – was found to be significantly different at the .001 level between those people with mental disorders and normal people. Then the TMHQ Profile was developed by using the Standard T-score.

Thailand is faced with the problems of rapid social and cultural changes that are responsible for

increasing mental health problems. Meanwhile, the proportion of mental health professionals to the

population is inadequate. So, a future mental-health policy must be well thought out. Because mental-health

treatment in the hospitals is neither sufficient nor effective, health promotion and education in the community

must be considered in order to find ways to prevent mental illness before professional treatment is required.

Moreover, psychiatric instruments which can be easily used in the community must also be considered.

The use of rating scales to detect mental disease is useful. The standard in psychiatric practice is usually a

class of persons, such as psychotic, neurotic, normal or other reference group. Many clinicians have found

that a schedule of items covering a variety of observable symptoms and relevant questions concerning

patients’ attitudes, feelings, and behaviors helps to assure a more thorough and complete psychiatric

interview or examination. There is less likelihood that possibly significant phenomena may be omitted or

overlooked. This is useful, whether or not the information so derived is later reduced to numerical form for

another purpose. Also, ratings are more likely to be interpreted within the same semantic frame of reference.

Clinicians with different backgrounds and holding different theoretical views find it valuable to have on

record information gathered in a systematic and common format^17,18. In addition, rating scales are very

standard norm profile. The result will be a more direct and meaningful application of an instrument to detect the mental health illness in the Thai community with a Thai standard norm profile.

based on the Diagnostic Criteria from the American Psychiatric Association’s Diagnostic and

completed TMHQ was printed. This included the test administration, test items, scoring summary,

from the total test items which was conducted from the information of the symptoms which have the

criteria from the DSM-IV. So, total sample size of this study came from the total test items multiplied by a factor of five which was the effective sample size to analyze the multivariate statistic (Factor Analysis) that was used in this study¹⁹. Therefore, the total samples in this study were 700 (normal samples = 350; psychiatric patient samples = 350). The sample was stratified on the variables of sex, age, occupation, and education.

of almost any symptom or group of symptoms of psychopathology. It was based upon a system analytic model of the DSM-IV. It was composed of 70 items. Each item is directed to indicate the extent to which the sample has been bothered or distressed by the problems or complaints represented by each of the 140 items, over a specified time interval, usually a one month period. The degree of distress ranges from “Not at all” to “Extremely”.

the results the TMHQ norm profile. All the data were analyzed by using SPSS for Windows. Subjects comprised 700 samples from two main groups: 350 normals and 350 psychiatric patients, includes 321 males and 379 females age between 15 to 60 years. The samples distributed by sex, age, occupation, income, and education.

Somatizarion: This dimension reflects distress arising from perceptions of bodily dysfunction. Complaints focus on cardiovascular, gastrointestinal, respiratory, and other symptoms with strong autonomic mediation. Headaches, pains, and discomfort localized in the gross musculature are also components, as are other somatic equivalents of anxiety.

Depression: Symptoms of dysphoric affect and mood are represented, as are signs of withdrawal of interest in life events, lack of motivation and loss of vital energy. The dimension mirrors feelings or hopelessness, worthlessness, meaninglessness, pessimism, loneliness, downheartedness, or discouragement. Several items are included concerning thoughts of death and suicidal ideation.

Anxiety: General indications such as restlessness, nervousness, and tension are included as are additional somatic signs. Items measuring free-floating anxiety, and panic attacks are an integral aspect of this dimension.

Psychotic: Florid, acute symptomatology, as well as behaviors typically viewed as more oblique, less definitive indicators of psychotic process are represented. In addition, secondary signs of psychotic behavior and indications of a schizoid life style are also represented.

Social function: General indications in interpersonal relationships and contact with other people in the social.

3. Administration. The standard instructions of the person at the top of the scale, and the items language, are in very basic Thai, so literacy level is rarely a problems.

4. Scoring and interpretation. Special answer sheets have been developed that can be read and scored. Each of the 5 scoring categories consists of from 10 to 20 items. By transferring the item responses to a separate scoring sheet the responses that make up a scoring category are added, and then divided by the number of items in that category. As a result, the score on any category can have a value ranging from 0, indicating no problem at all, to a maximum of 4, when there is an extreme problem (this is for the negative items, but for the positive items, the score is conversed). And, since all categories are interpreted the same way, screening of even a large number of examinees’ records can be quickly accomplished just by scanning the raw scores.

5. Scoring summary. A summary score is a quantitative index of the degree to which a particular problem area is relevant to a client. The TMHQ’s scoring summary is as follows:

The TMHQ’s Profile was created in an easily construct profile. The user can visualize it because it is translated numerical values into a plotted profile.

Table 2 presents mean, standard deviation, and t-test between the high and low groups.

Table 3, 4 and 5 showed the scoring summary of the TMHQ, Factorized Score, and The TMHQ Norm Profile.

High group (n= 350)		Low group (n= 350)		High group (n= 350)		Low group (n= 350)
M	SD	M	SD	M	SD	M	SD
1	2.21	.74	3.26	.74	-17.50*	36	3.36	.59	2.31	.63	8.00*
2	3.13	.74	2.15	.79	14.00*	37	2.35	.88	3.34	.63	21.00*
3	3.00	.64	1.94	.89	15.14*	38	3.06	.56	2.14	.80	14.14*
4	2.98	.68	2.08	.75	15.00*	39	2.20	.74	3.06	.66	-15.33*
5	3.27	.61	2.17	.93	18.33*	40	2.88	.61	2.14	1.01	14.33*
6	3.27	.52	2.27	.85	16.67*	41	2.19	.75	3.48	.78	10.57*
7	3.14	.69	2.31	.81	11.86*	42	2.02	.80	3.25	.73	18.43*
8	3.16	.62	2.27	.67	17.80*	43	3.48	.65	2.10	.83	17.57*
9	3.10	.86	2.11	.73	14.14*	44	2.18	.67	2.87	.85	-19.71*
10	2.25	.91	3.43	.30	-19.67*	45	2.19	.69	2.80	.75	-10.17*
11	3.59	1.50	1.98	.83	16.10*	46	2.75	.85	2.19	.76	8.00*
12	2.32	.73	1.17	.76	19.17*	47	3.05	.87	2.12	.86	13.29*
13	2.89	1.16	2.18	.76	8.88*	48	3.21	.73	2.09	.84	16.00*
14	3.06	.75	2.23	.78	11.86*	49	3.17	.92	2.19	.76	14.00*
15	3.18	.67	2.18	.68	16.67*	50	2.70	.63	1.93	1.06	13.86*
16	3.09	.76	2.06	.86	14.71*	51	3.08	.84	2.18	.76	12.86*
17	3.27	.61	2.32	.73	19.00*	52	2.65	.63	1.52	.88	16.14*
18	3.40	.74	2.10	.83	18.57*	53	2.89	.82	1.92	1.01	12.13*
19	3.00	.58	2.21	.74	15.80*	54	3.25	.59	2.18	.76	17.83*
20	2.92	.78	1.18	.73	12.33*	55	2.28	.79	3.17	1.59	-8.90*
21	3.24	.62	2.09	.84	16.43*	56	1.77	.99	2.70	.99	-7.75*
22	2.81	.77	1.05	.69	12.67*	57	2.35	.88	3.34	.63	21.00*
23	2.98	.82	2.74	.69	4.00*	58	2.31	.63	3.06	.76	-12.50*
24	2.95	.77	2.01	.81	13.43*	59	3.06	.66	2.20	.74	15.33*
25	2.62	.68	2.02	.80	10.00*	60	2.92	.78	1.18	.73	12.33*
26	3.48	.50	2.24	.78	20.67*	61	2.18	.75	3.48	.78	10.57*
27	3.02	.80	2.16	.78	21.67*	62	2.81	.77	1.05	.69	12.67*
28	3.36	.61	2.28	.66	12.29*	63	2.18	.76	2.89	1.16	-8.88*
29	3.27	.54	2.07	.77	21.60*	64	2.75	.75	2.98	.68	-15.00*
30	3.16	.63	2.17	.76	20.00*	65	2.06	1.00	2.62	.72	-6.83*
31	3.41	.63	2.08	.92	16.33*	66	2.75	.85	2.19	.76	-12.50*
32	3.25	.86	2.23	.89	22.17*	67	2.31	.81	3.14	.69	-11.86*
33	2.58	.96	2.24	.70	12.88*	68	1.18	.73	2.92	.78	-12.33*
34	2.63	.63	2.22	.81	4.71*	69	2.14	1.01	2.88	.61	-14.33*
35	2.62	.72	2.06	1.00	6.83*	70	2.22	.81	2.63	.63	-4.71*

All the TMHQ items exhibited a reasonable spread of responses to discriminate between those with mental disorders from normal people. In the assessment of psychological distress, a primary indicator of mental health should be discriminated between the mental disorders from normal people². In this study, all of the TMHQ items were found to be significantly different between the high and low responses group. Therefore, the 70 items of the TMHQ met the criteria for discriminant validity.

The TMHQ was designed to assess a domain of functioning. It is factor-analyzed to identify separable dimensions, representing theoretical constructs, within the domain. Psychometricians strongly recommend that test developers should begin by factor analyzing the items^3,4,5,6.

The TMHQ Items. The principal components analysis was carried out 70 items. Therefore, the final TMHQ is composed of 70 items. The number of items in a rating scale can vary from one (in particular global scales) to several hundred (such as MMPI). Investigations with intent structure analysis have demonstrated that 5-10 items are usually sufficient when the total score of the scale is utilized as a measurement of the final rating or evaluation¹⁶. According to Bech et al., the number of items in the TMHQ are sufficient to measure multidimensional psychopathology.

Administration. The standard instructions of the person at the top of the scale, and the items language, are in very basic Thai, so literacy level is rarely a problems. Use of the scale is simple enough so that the examinee does not usually need any more explanations. That is the basic principle of item writing²⁰. The examinee is directed to indicate the extent to which he/she has been bothered or distressed by the problems or complaints represented by each of the 62 items over a specified time interval, usually a one month period. That is the same as other symptom scales as screening or diagnosis scales such as PSE (Present State Examination), GHQ (General Health Questionnaire), and SCL-90 (Symptom Distress Check-List), for example. It normally takes about 15 to 20 minutes to complete the TMHQ’s scale.

Scoring and interpretation. The TMHQ scores can be interpreted in three ways. First, they can be regarded as a measure of the severity of psychological disorder. Second, they can be used to estimate the prevalence of psychiatric illness. Third, they can be regarded as an indicator of morbidity.

The TMHQ Profiles: To characterize each item in the symptom scale by means of a quantitative index, the five scales are: No stress reported, Some stress but infrequent and of low intensity, Somewhat regular stress of mild or moderate intensity, Regular stress of moderate to high intensity, and Examinees experiences extreme stress associated with these symptoms due to frequency, intensity or both. Likert ²⁰ found that five degree categories were optimal, and Fridenberg ²¹supported this. The TMHQ’s Profile was created in an easily construct profile. The user can visualize it because it is translated numerical values into a plotted profile.

The limitation of the TMHQ for diagnosis purposes. The TMHQ nomothetic indices may be appropriate for screening purposes or to indicate areas in need of further assessment. However, it does not provide sufficiently specific information concerning the parameters of situational determinants of the problem behavior. Therefore, other behavioral assessment methods (e.g., interviews, self-monitoring, observational procedures) also contribute to the assessment function. Questionnaires could be of more value to target problem description and specification than is using only one measurement. Moreover, increased emphasis on issues of the multimodel nature of behavior problems would be beneficial in clinical¹³.

The Thai Mental Health Questionnaire has sufficient discriminant power, adequate construct

validity, adequate reliability, ant the Norm Profile. The result will be a more direct and meaningful application of an instrument to detect the mental health illness in the Thai community.

This study was supported by Chalermprakriet Fund 1998. Faculty of Medicine, Siriraj Hospital. The authors would like to thank Professor Dr. Nancy Jones for her advices on

3. Briggs SR, Cheek JM. The role of factor analysis in the development and evaluation of

4. Comrey AL. Factor analysis methods of scale development in personality and clinical psychology.

5. Cortina JM. What is Coefficient alpha? An Introduction of theory and applications.

6. Floyd FJ, Widaman KF. Factor analysis in the development and refinement of clinical

8. Cole DA. Utility of Confirmatory factor analysis in test validation research. J Consult

9. Hayduk LA. Structural equation modeling with LISREL: Essentials and advances. Baltimore:

10. Bollen KA. Structural equations with latent variances. New York: Wiley, 1989.

11. Kenny DA, Kashy DA. Analysis of the multitrait-multimethod matrix by confirmatory factor analysis. Psychol Bull 1992;112:165-72.

12. Streiner DL. Figuring out factors: The use and misuses of factor analysis. Can J Psychiatry 1994; 39:135-40.

13. Jansen B, Haynes S. Self-report Questionnaires and Inventories. Handbook of Behavioral Assessment (3rd ed.). New York: John Wiley & Sons, 1989.

14. Fischer F, Corcoran K. Measures for Clinical Practice: A Source Book, Volumn 1 (2nd ed).

15. Aiken LR. Psychological Testing and Assessment (2nd ed.). Boston: Allyn and Bacon. 1994.

16. Bech P, Malt UF, Dencker SJ. Scales for Assessment of Diagnosis and Severity of Mental Disorders.

17. Staff of Research and Education Association. Handbook of Psychiatric Rating Scales. New York:

18. MacKinnon RA, Yudofsky SC. Principles of the Psychiatric Evaluation. New York: Lippincott

19. Goldberg DP, Williams PA. User’s Guide to the General Health Questionnaire. London:

20. Likert RA. technique for the measurement of attitudes. Archives Psychological 1932;4:4.

21. Fridenbert M, Carroll BJ. The Carroll Rating Scale for Depression. In: Sartorius N, Ban T A, eds.

22. Clark LA, Watson D. Constructing Validity: Basic Issure in Objective Scale Development.

พัฒนาโดย ผ.ศ. ดร. สุชีรา ภัทรายุตวรรตน์ และคณะ ภาควิชาจิตเวชศาสตร์ คณะแพทยศาสตร์ศิริราชพยาบาล มหาวิทยาลัยมหิดล