Cross-sectional study of General Health Questionnaire among a University Students in
Malaysia: A Reliability Study.
By
Dr. Prashanth Talwar1 and Prof. Mohd Fadzil Abdul Rahman2
Dr. Prashanth Talwar1Ph.D
Associate Professor
Universiti Malaysia Sarawak
94300 Kota Samrahan.
talpra@rediffmail.com
Prof. Mohd Fadzil Abdul Rahman2 MA
Professor and Deputy Vice Chancellor
Universiti Malaysia Sarawak
94300 Kota Samrahan.
Corresponding Author
Dr. Prashanth Talwar1Ph.D
Associate Professor
Universiti Malaysia Sarawak
94300 Kota Samrahan.
talpra@rediffmail.com
Statistical summary of the manuscript
Total Number of words: 3984
Number of words in abstract: 161
Number of references: 32
Number of Tables: 1
Cross-sectional study of General Health Questionnaire among a University Students in
Malaysia: A Reliability Study.
Abstract
Background: The General Health Questionnaire is a widely used scale to measure
psychological distress. This scale is available in many languages. The original version
comprised of 60 items. It is available in 30-items, 28-items, 20-items and 12-items. The
shorter version is as good as the longer version of the scale. Objective: The aim of this
investigation was to examine the psychometric properties and reliability of the 12-item
General Health Questionnaire (GHQ-12) among university students. Method: The study was
conducted in one of universities in Malaysia. A total of 280 students were selected using
convenient sampling. Result: The results of this study showed that the GHQ-12 is
multidimensional and contains three factor structures. It has a high internal consistency and a
useful instrument to measure the psychological wellbeing of university students in Malaysia.
Conclusion: Although several studies have been conducted across the globe on the GHQ-12,
debate surrounds in relation to its dimensionality, negative and positive wordings and cut off
point.
Key Words: GHQ Factor Structure Alpha Reliability.
1
Cross-sectional study of General Health Questionnaire among a University Students in
Malaysia: A Reliability Study.
Introduction:
It goes without saying that life in the university is stressful. Stress arises due to a variety of
factors such as social adjustment to the environment, assignments, peer relationship, peer
pressure to achieve a high score in exams. The ability to adjust to stress depends on the
individual coping strategies. Over exposure to stress can cause physical, emotional and
mental health problems 1. In study done by Zaid et al 2 in Malaysia, it was found that the
prevalence of emotional disorders among students was high. Zaid et al 2 reported that there
was significant association between emotional disorders and respondents' relationship with
their parents, siblings and lecturers, as well as level of pressure prior to exam. It is important
to detect emotional disorders at an early stage so that treatment necessary can be given to
those affected 2.
Psychological distress can be measured using standardized instruments, which can help in
detecting cases who have psychological distress. The General Health Questionnaire (GHQ12) is a self-reporting measure. This scale focuses on breaks in normal functioning rather
than on life-long traits; therefore, it only covers disorders or patterns of adjustment associated
with distress 4. The original version is composed of 60 items. In recent years the 12-item
General Health Questionnaire (GHQ-12) has been extensively used as a short screening
instrument, producing results that are comparable to longer versions of the GHQ 5. The GHQ12 is a measure of current mental health. It focuses on two major areas- inability to carry out
normal functions and the appearance of new and distressing experiences 4.
The 12-item General Health Questionnaire (GHQ-12) has been widely used in many
countries for detecting psychological morbidity 6. It has been translated and validated in
various languages 5. The GHQ-12 is also available in Malay language where the internal
consistency was excellent 7. A high degree of internal consistency was observed for each of
the 12 items in the Malay version 8.
If investigators wish to use a screening instrument as a case detector, the shorter GHQ is
remarkably robust and works as well as the longer instrument. The latter should only be
preferred if there is an interest in the scaled scores provided in addition to the total score5.
The GHQ threshold is partly determined by the prevalence of multiple diagnoses, with higher
thresholds being associated by higher rates of both single and multiple diagnoses. The mean
GHQ score for the whole population of respondents provides a rough guide to the best
threshold 9.
The GHQ-12 was designed as a uni-dimensional scale with positively phrased and negatively
phrased items. Many factor-analytic studies have reported that the GHQ-12 has two or three
dimensions, threatening its validity 10.
Reviewing past research which indicates that the GHQ-12 is a brief and easy to understand
screening instrument it was intended to investigation the psychometric properties of the
General Health Questionnaire (GHQ-12) among university students. The aim of this
2
investigation was to examine the psychometric properties and reliability of the 12-item
General Health Questionnaire (GHQ-12) among university students.
Materials and Methods
Participants
The study was conducted in one of the universities in Malaysia. Please be specific which
university? The study participants were selected using convenience sampling. The
questionnaire consisting of part 1 which included the demographic aspects and part 2, the
General Health Questionnaire (GHQ-12). Permission was obtained from the university before
conducting this study. Undergraduate students were told about the purpose of the study and
only those volunteered to participate were given the self-administered questionnaire in the
class room. Data was collected from 300 students. Seven percent% of the questionnaire had
to be excluded due to incomplete information. The data obtained from 280 students was
taken up for final data analysis.
Instrument
Goldberg’s General Health Questionnaire (GHQ-12) with 12 items has four responses. The
GHQ-12 can be classified as either positively worded or negatively worded. Six items
referred to health is considered to be positively worded. The response range from ‘more than
usual’ to ‘much less than usual’. 6 items referring to disease are negatively worded. The
response ranges from ‘not at all’ to ‘much more than usual’. Items can be scores according to
the three scoring methods. Likert method (all items coded 0-1-2-3), GHQ method (all items
coded 0-0-1-1), and C-GHQ method (PP items coded 0-0-1-1; NP items coded 0-1-1-1).
In this study, the Likert method of scoring (0-1-2-3) was chosen. The scores were summed up
by adding all the items on the scale ranging from 0 to 12. Due to the various thresholds of the
GHQ-12, the mean GHQ score for a population of respondents was suggested as a rough
indicator for the best cut-off point 9. Based on the mean score from the study the cut-off point
12 was used. The least one can score is 0 and the maximum one can score is 36.
Statistical analysis
SPSS 21 was used to process the data. Descriptive analysis was done to study the frequency
of students who with least score and highest score. Dimensionality was assessed using factor
analysis. Exploratory factor analysis (EFA) was done using the principle component analysis.
In this study the variance explained by the factor structure of the GHQ-12 for one-factor,
two-factor and three-factor solutions was assessed.
Rotation method used was Varimax with Kaiser Normalization. Factor loading was based on
absolute value greater than 0.40. Cronbach’s11alpha coefficient was used to calculate the
reliability of the scale. Cronbach's alpha is an index of reliability associated with the variation
accounted for by the true score of the "underlying construct." Construct is the hypothetical
variable that is being measured 12. The higher the score, the more reliable the generated scale
is. Nunnaly 13 has indicated 0.7 to be an acceptable reliability coefficient.
Results
The mean age of the respondents was 21 years. 34 % were male and 66% were female
student respondents.
The mean GHQ-12 score was 12.32 (SD=5.23). The least one has scored was 3 and
maximum score was 23. Taking 12 to be the cut-off point about 53% of the respondents
scored more than or equal to 12.
3
Factor Structure
KMO value is 0.813, which is considered meritorious. Results of Bartlett’s test of sphericity
shows 2=1163.07 (df=66, p<0.000). In this study, the sample inter correlation matrix did not
come from a population in which the inter correlation matrix is an identical matrix. There was
no correlation error among the variables.
Based on eigen value more than 1, results from EFA revealed three dimensional structures.
The first factor includes items 3,4,5,6,8 and 9 and explains 22.65 % variance. The second
factor includes items 7, 10, 11 and 12 and explains 19.84 % variance. The third factor
includes two items 1 and 2 and explains 16.53 % variance. Total variance explained by the
three factors is 59.03 %.
Reliability.
The internal consistency was measured using Cronbach's alpha. The overall Cronbach's alpha
for the entire sample in this study was 0.84. Alpha value for male students was 0.85 and that
for the female students was 0.83 respectively.
Table No. 1 shows correlation between items and Cronbach's alpha if Item were to be
deleted. The overall Cronbach's alpha was 0.84. Only item no 1 “Been able to concentrate on
whatever you are doing?” showed the lowest correlation coefficient of 0.32. All the
remaining items showed a correlation ranging from 0.45 to 0.59. Eliminating item No. 1 did
not change the value of Cronbach's alpha substantially.
Table No 1
Correlation between items and Reliability analysis of GHQ-12.
No.
Item
Corrected
Cronbach's
Item-Total Alpha if
Correlation Item
Deleted
Been able to concentrate on whatever you are
GHQ1
.321
.838
doing?
Lost much sleep over worry?
GHQ2
.527
.823
Felt that your were playing a useful part in
GHQ3
.533
.823
things?
Felt capable of making decisions about things?
GHQ4
.456
.829
Felt constantly under strain?
GHQ5
.549
.822
Felt that you could not overcome your
GHQ6
.595
.818
difficulties?
Been able to enjoy your normal day-to-day
GHQ7
.453
.829
activities?
Been able to face up your problems?
GHQ8
.472
.828
Been feeling unhappy and depressed?
GHQ9
.467
.828
Been
losing
confidence
in
yourself?
GHQ10
.501
.825
Been
thinking
of
yourself
as
a
worthy
person?
GHQ11
.582
.819
Been feeling reasonably happy, all things
GHQ12 considered?
.545
.822
4
Discussion
It is necessary for students to be in their good and balanced psychological health in order to
excel in their pursuit and for a successful future by contributing positively towards human
capital resources of the country 14. Stress among university students in Malaysia is high. In a
study done by medical students in Malaysia it was found a total of 41.9% of the medical
students were found to have emotional disorders 15.
The GHQ-12 was used to assess the psychological wellbeing of university students. The
GHQ-12 item questionnaire is the most extensive (are you sure this is the most extensive ?
even as compared to long version?) used screening instrument for common mental disorders,
in addition to being a more general measure of psychiatric well-being16. The aim of this
investigation was to examine the psychometric properties and reliability of the 12-item
General Health Questionnaire (GHQ-12) among university students.
The mean GHQ-12 score was 12.32 (SD=5.23). A cut off point of 12 or higher was
considered as those having problems. Taking 12 to be the cut-off point based on the mean
about 53% of the respondents scored ≥12. This shows that a sizeable portion of students
experienced psychological distress. The least score was 3 and maximum score was 23.
Using Kaiser-Meyer Olkin test of Sphercity it was found the GHQ-12 obtained from this
sample was suitable to conduct factor analysis and sample size was appropriate using the
Bartlett’s test of sphericity. Agree with the statement.
Different versions of the General Health Questionnaire (GHQ), including the GHQ-12 and
GHQ-28 have been subjected to factor analysis in a variety of countries17. In this study,
exploratory Factor analysis of GHQ-12 revealed three dimensional structures. Factor I
consisted of 6 items explained 22.65 % variance. Factor II consisted of 4 items explained
19.84 % variance. Factor III consisted of 2 items explained 16.53 % variance. Total variance
explained by the three factors is 59.03 %. It was found that the loadings ranged from 0.46 to
0.80. Majority of the loading was >0.50. The number of items loaded differed from one study
to another. This study is similar to the study done by Sánchez-López 16 who reported a three
factor model with 54.19 % explained variance. This study is also consistent with studies done
in Malaysia by Zulkefly & Baharudin 14 which explained three factor structure with 51.9%
variance and fits into the three factors namely psychological distress, social and emotional
dysfunction and cognitive disorder, thought the items may slightly differ from one another.
Unidimensional or multidimensional – The GHQ-12 was designed as a uni-dimensional scale
with six items positively phrased and remaining six items negatively phrased items. Few
studies have reported GHQ-12 to be uni-dimensional, other studies reported two dimensions
and yet few others, three dimensional 6, 18.
The factorial structure of GHQ-12 may be debatable in view of discrepancy in reported by
authors. Validation studies have assumed that the GHQ-12 is one-dimensional and free of
response bias, but recent evidence suggests that neither of these assumptions may be correct,
threatening its utility as a screening instrument19.
Like any other instruments GHQ-12 too comes with its own limitations. Three aspects of
GHQ-12 is contentious, the first being the number of dimensions - which would be the best
fit; the second is the cut-off point and third is the positive and negative wordings. In this
study attempts are made to debate the factor structures however, to avoid lengthy discussions
only two or more authors work has been cited for comparison.
5
Single Factor Model. Using the Likert-type scoring approach, Hankins's one-factor model
with "method effects" obtained the best fit20. Confirmatory factor analyses showed that the
one-factor structure, controlling by the wording effect, gathered the best fit indexes. It was
concluded that the one-factor model of the GHQ-12 is more appropriate21. Studies done by
Lesage 22 supports a single factor structure, as in the original Goldberg’s version. However,
according to Gao et al 6 the one-dimensional model was the worst according to all three
goodness-of-fit indices.
Two Factor Model. The principal component analysis with oblique rotation solution showed
that the GHQ-12 was a measure of psychological morbidity with two-factor structure, that
jointly accounted for 51% of the variance23. In another study, Hu et al 24 reported that the best
description of the factor structure of the GHQ-12 is given by a variant of a two-factor
solution, corresponding to positive mental health and symptoms of mental disorder. Kalliath
et al 25 reported that the results favour a two-factor model consisting of a ‘Social
Dysfunction’ factor and an ‘Anxiety/Depression’ factor measured by four items each.
Three Factor Model.
The CFA of the GHQ-12 indicated that the best fit and the simplest solution were provided
by the three-factor solution in both subpopulations 26. Graetz 27 examined the factor structure
of the 12 item GHQ and proposed that the model with the best fit is the three factor model.
Several studies have found that Graetz's 3-factor model of the GHQ-12 is more plausible than
other models 28. In a study done by Padrón 29, the model with the best fit in the CFA was the
three-dimensional model, social dysfunction, anxiety and self-esteem, the three factors
together explained 53.7% of the variance.
It is not necessary that the three factor model is the best solution. Salama-Younes et al
suggests that analysis for the GHQ-12 demonstrated a good fit not only to the two-factor
model (positive vs. negative items) but also to a three-factor model. Further, Padrón 29 also
reported a good fit in two factor as well as three factor model.
30
In this study the variance explained by the factor structure of the GHQ-12 for one-factor,
two-factor and three-factor solutions was assessed, it was found that one factor explained
36.48 % variance, two factor explained 49.47 % variance and the three factor solution
explained 59.03 % variance. Therefore, it is sensible to accept the three factor model as the
best solution.
The overall Cronbach's Alpha for the entire sample in this study was 0.84. In a similar study
done by Yusoff et al 31. Cronbach’s alpha values for the GHQ-12 was of 0.85. Thus, this
study matches with study done elsewhere in Malaysia. Study also matches with studies done
in Iran, where Montazeri et al 23, found Cronbach's alpha coefficient of 0.87. Alpha value for
male students was 0.85 and that for the female students was 0.83 respectively. This shows a
good fitting of internal consistency for both the groups.
Anther debate that surrounds the GHQ-12 is the cut-off point. The cut-off point for GHQ-12
varies from one study to another. The finding of variation in the optimal threshold of the
General Health Questionnaire (GHQ) across different settings has proved difficult to explain
32
. No convincing explanation has been forthcoming for the variation in best threshold to
adopt for the GHQ in different settings 9. According to Godlberg et al 9, the GHQ threshold is
partly determined by the prevalence of multiple diagnoses, with higher thresholds being
6
associated by higher rates of both single and multiple diagnosis. The mean GHQ score for the
whole population of respondents provides a rough guide to the best threshold 9. In this study
the cut-off was arrived at based on the mean.
Positively worded and negatively worded items too could be a problem. According to
Hankins, conventional psychometric assessments using factor analysis and reliability
estimates have obscured substantial measurement error in the GHQ-12 due to response bias
on the negative items, which limits its utility as a screening instrument for psychiatric
morbidity 19. Hu et al 24, is of the opinion that, one weakness lies in the GHQ-12 itself. All
self-report questionnaires are prone to method variance, namely the tendency for people to
respond the same way to similarly worded items. This may contribute to the aggregation of
responses to positively and negatively worded items. Although statistical methods exist for
distinguishing between this and ‘true’ variance due to latent constructs, these methods are not
applicable where there are only one or two construct factors 24.
In conclusion, the GHQ-12 is a widely used scale to measure psychological health. This scale
has been tested and validated in many languages. As found in this study other studies too
show that this scale has a good internal consistency. There is difference of opinion regarding
the factor structure of the scale as some authors say that one factor model is the best fit others
do not agree with one factor model and suggest two or three factor model. Others have shown
that the two factor model is as good as three factor model. There has also been difference of
opinion regarding the cut-off point, as different authors suggesting different cut-off point.
However, the cut-off point is based on the mean was suggested by the author of GHQ.
Probably it will be good to find the mean and best cut-off point for the Malaysian population.
There is also some debate surrounding the wording of GHQ-12. Few authors feel that
negative worded items cause error in measurement. It may be suggested that further studies
need to be done taking into consideration arguable questions.
7
REFERENCES
1.
Niemi PM & Vainiomäki P. T. Medical Students’ Academic Distress, Coping, and
Achievement Strategies During the Preclinical Years. Teaching and Learning in
Medicine. 1999;11:125-34.
2.
Zaid ZA, Chan SC & Ho JJ. Emotional disorders among medical students in a
Malaysian private medical school. Singapore Medical Journal. 2007; 48(10): 895–899.
3.
Sherina M, Rampal L, & Kaneson N. Prevalence of emotional disorders among medical
students in a Malaysian university. Asia Pacific Family Medicine. 2003; 2(4): 213–217.
4.
Goldberg D & Williams P. A user’s guide to the General Health Questionnaire. Slough:
NFER-Nelson; 1988.
5.
Goldberg DP, Gater R, Sartorius N , Ustun TB, Piccinelli M , Gureje O , & Rutter C.
The validity of two versions of the GHQ in the WHO study of mental illness in general
health care. Psychological Medicine. 1997; 27(1): 191–197.
6.
Gao F, Luo N , Thumboo J , Fones C, Li SC, & Cheung YB. Does the 12-item
General Health Questionnaire contain multiple factors and do we need them? Health
and Quality of Life Outcomes. 2004; 2: 63.
7.
Saiful M BY . The validity of two Malay version of the General Health Questionnaire
(GHQ) in detecting distressed medical students. ASEAN Journal of Psychiatry.
2010;11.Retrived from http://www.aseanjournalofpsychiatry/pdf/ASEAN_/10201.pdf
8.
Quek KF, Low WY, Razack AH, & Loh CS. Reliability and validity of the General
Health Questionnaire (GHQ-12) among urological patients: A Malaysian study.
Psychiatry and Clinical Neurosciences . 2001; 55(5): 509–513.
9.
Goldberg DP , Oldehinkel T, & Ormel J. Why GHQ threshold varies from one place to
another. Psychological Medicine. 1998; 28(4): 915–921.
10.
Hankins M. The factor structure of the twelve item General Health Questionnaire
(GHQ-12): the result of negative phrasing? Clinical Practice and Epidemiology in
Mental Health. 2008; 4: 10.Retrived from
http://www.cpementalhealth.com/content/4/1/10.
11.
Cronbach LJ. Coefficent alpha and the internal structure of tests. Psychometrika. 1951;
16: 297-334.
12.
Hatcher L. A step-by-step approach to using the SAS(R) system for factor analysis and
structural equation modelling. Cary. NC: SAS Institute; 1994.
13.
Nunnaly J. Psychometric theory. New York: McGraw Hill; 1978.
14.
Zulkefly N S, & Baharudin R. Using the 12-item General Health Questionnaire ( GHQ8
12 ) to Assess the Psychological Health of Malaysian College Students. Global Journal
of Health Science. 2010; 2(1):73–80.
15.
Salam A, Yousuf R, Bakar SMA, Haque M. Stress among Medical Students in
Malaysia: A Systematic Review of Literatures. International Medical Journal. 2013; 20
(6): 649 – 655.
16.
Sánchez-López MDP, & Dresch V. The 12-Item General Health Questionnaire (GHQ12): reliability, external validity and factor structure in the Spanish population.
Psicothema. 2008; 20(4): 839–43. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/18940092
17.
Werneke U, Goldberg D P, Yalcin I, & Ustün B T. The stability of the factor structure
of the General Health Questionnaire. Psychological Medicine. 2000; 30(4): 823–829.
18.
Shevlin M & Adamson G. Alternative factor models and factorial invariance of the
GHQ-12: a large sample analysis using confirmatory factor analysis. Psychological
Assessment. 2005; 17(2):231–236.
19.
Hankins M. The reliability of the twelve-item general health questionnaire (GHQ-12)
under realistic assumptions. BMC Public Health. 2008; 8: 355.Retrived from
http://biomedcentral.com/1471-2458/8/355.
20.
Aguado J, Campbell A, Ascaso C, Navarro P, Garcia-Esteve L & Luciano JV.
Examining the Factor Structure and Discriminant Validity of the 12-Item General
Health Questionnaire (GHQ-12) Among Spanish Postpartum Women. 2008;
Assessment. 2012; 19(4):517-25.
21.
Gouveia VV. General Health Questionnaire ( GHQ-12 ): the effect of negative items in
its factorial structure . Cad Saude Publica.2012; 28(2):375–384. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/22331163.
22.
Lesage FX. Validation of the General Health Questionnaire (GHQ-12) adapted to a
work-related context. Open Journal of Preventive Medicine. 2011; 01(02): 44–48.
23.
Montazeri A , Harirchi, AM, Shariati M, Garmaroudi G, Ebadi M, & Fateh A. The 12item General Health Questionnaire (GHQ-12): translation and validation study of the
Iranian version. Health and Quality of Life Outcomes. 2003;1: 66.
24.
Hu Y, Stewart-Brown S, Twigg L , & Weich S. Can the 12-item General Health
Questionnaire be used to measure positive mental health? Psychological Medicine.
2007; 37(7): 1005–13.
25.
Kalliath T J, O’Driscoll MP, & Brough P. A confirmatory factor analysis of the General
Health Questionnaire-12. Stress and Health. 2004; 20(1): 11–20.
26.
Penninkilampi-Kerola , Miettunen J, Ebeling H. A comparative assessment of the factor
structures and psychometric properties of the GHQ-12 and the GHQ-20 based on data
from a Finnish population-based sample. Scandinavian Journal of Psychology. 2006;
9
47(5):431-40.
27.
Graetz B. Multidimensional properties of the General Health Questionnaire. Social
Psychiatry and Psychiatric Epidemiology. 2004;26(3):132–138.
28.
Gao W, Stark D, Bennett, MI, Siegert R J, Murray S & Higginson I J. Using the 12item General Health Questionnaire to screen psychological distress from survivorship
to end-of-life care: Dimensionality and item quality. Psycho-Oncology. 2012; 21(9):
954–961.
29.
Padrón A, Galán I, Durbán M, Gandarillas A, & Rodríguez-Artalejo F. Confirmatory
factor analysis of the General Health Questionnaire (GHQ-12) in Spanish adolescents.
Quality of Life Research : An International Journal of Quality of Life Aspects of
Treatment, Care and Rehabilitation. 2012; 21(7): 1291–8.
30.
Salama-Younes M, Montazeri A, Ismaïl A & Roncin C. Factor structure and internal
consistency of the 12-item General Health Questionnaire (GHQ-12) and the Subjective
Vitality Scale (VS), and the relationship between them: a study from France. Health
and Quality of Life Outcomes. 2009;7: 22. Retrived from
http://www.hqlo.com/content/7/1/22
31.
Yusoff M SB, Abdul Rahim A F & Yaacob M J. Prevalence and Sources of Stress
among Universiti Sains Malaysia Medical Students. The Malaysian Journal of Medical
Sciences :2010; 17(1): 30–7. Retrieved from
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3216143&tool=pmcentrez
&rendertype
32.
Willmott S A, Boardman JAP, Henshaw C A & Jones PW. Understanding General
Health Questionnaire (GHQ-28) score and its threshold. Social Psychiatry and
Psychiatric Epidemiology. 2004; 39(8): 613–617.
The authors failed to address why there is a need to revalidate the GHQ as previous
studies were conducted to look at the psychometric properties of this scale. What is the
strength of this study as compared to other studies? In the result there is no validity
analysis such as the analysis of factor loading, however there is a discussion on factor
models. Please describe the detail of handling data missing of 20 subjects which were
excluded in the study. Convenient sampling is one of the setbacks of the study which
could be highlighted. Please display the demographic profiles of your subjects as the
level of education and races may influence the psychometric properties of a scale.
My conclusion is the manuscript is publishable with several modifications.
10