The Accuracy of Indonesian Version of HAM-A

Juliana Irmayanti Saragih

and Etti Rahmawati

Department of Clinical Psycology, Universitas Sumatera Utara, Jl. Dr. Mansyur No.7, Medan, Indonesia

Department of Experiment, Universitas Sumatera Utara, Jl. Dr. Mansyur No.7, Medan, Indonesia

Keywords: Test Anxiety, HAM-A, Hamilton Anxiety Scale.

Abstract: This study is part of a research to see if Diaphragmatic Breathing and Progressive Muscle Relaxation are

effective in reducing students' anxiety during tests. One tool that can be used to measure anxiety is the

Hamilton Anxiety Rating Scale (HAM-A). To see if HAM-A has accurate quality measurement results to

measure anxiety, it is tested upon the validity and reliability by involving 220 participants. The adaptation

process begins by examining the coexistence of anxiety constructs that exist in HAM-A. The next process is

an estimate of the correlation of HAM-A with the dimensions of the neurotic personality to prove convergent

validity. The Neurotical Subscale is one of five subscales in the Big Five Inventory (BFI). Based on the

analysis and review of the results, it can be concluded that the Hamilton Anxiety Rating Scale Scale (HAM-

A) which has been translated into the Indonesian language has an accurate quality measurement results to

measure anxiety.

1 INTRODUCTION

One of the factors that make student academic

performance is not optimal is anxiety and one of its

type is anxiety when facing examinations (Burns,

2004). Many students who experience anxiety during

the exam state that they are difficult to concentrate on

questions during the exam, which ultimately makes

their academic performance unsatisfactory even if

they know the material that is being tested (Amiri &

Ghoonsoly, 2015). This anxiety also appears to be

related to some other problems both physically and

psychologically in an academic setting. Students who

feel anxiety cannot optimize their potential while

undergoing the exam (Hancock, 2001). Anxious

students will also get lower test results (Everson et al.,

1991), more difficult to learn new materials in the

classroom and generally lower academic

performance (Chapell et al., 2005) They are also

reported to have low motivation, self-assessments

tend to be negative and difficult in concentration

(Swanson & Howell, 1996) even leading to suicidal

behavior (Schaefer et al., 2007).

The anxiety of facing a student exam can be

measured by the Hamilton Anxiety Rating Scale

(known as HAM-A) (Hamilton, 1959). Hamilton

Anxiety Rating Scale is a scale created by Max R

Hamilton in 1959 to measure anxiety levels

experienced by a person. The high level of student's

anxiety in facing the exam appears on the score

obtained on the scale of HAM-A. The higher the

score gained means the higher the anxiety felt by the

student when facing the exam, the lower the score

obtained means the lower the anxiety level

1.1 Validity

Validity refers to how far the evidence and theories

support the interpretation contained in the test scores

required for the use of the test itself. Validity is not

seen from the test instrument as a research instrument,

but viewed from the measurement results of a test

device (Osterlind, 2010).

Evidence of validity based on internal structure

emphasizes the study of the construction of

measuring instruments in accordance with the theory

on which it is based. With proof of internal structure,

we will be able to know whether a measuring

instrument is really a representation of the latent

attribute to be measured. Analysis model of validity

evidence with internal structure can be done with

Multitrait-multimethod and confirmatory factor

analysis. Evidence of validity based on the

relationship with other variables is a review of how

far the scores obtained are related to other scores

(criteria). Procedures that can be done through

Saragih, J. and Rahmawati, E.

The Accuracy of Indonesian Version of HAM-A.

DOI: 10.5220/0010088015451549

In Proceedings of the International Conference of Science, Technology, Engineering, Environmental and Ramiﬁcation Researches (ICOSTEERR 2018) - Research in Industry 4.0, pages

1545-1549

ISBN: 978-989-758-449-7

1545

experimental and correlational studies based on

nomological network construct are measured.

1.2 Item Discrimination

Ebel (Azwar, 2013) mentions that the evaluation of

the item discrimination power index can be divided

into four categories, namely: item with a

discrimination power index of 0.40 or more is

categorized as excellent, 0.30-0.39 is categorized as

good, but needs to be improved. While the

discrimination power index of 0.20-0.29 is

categorized as unsatisfactory or needs to be corrected

and the item discrimination power index of less than

0.29 is bad and should be discarded.

1.3 Reliability

Reliability refers to the accuracy of measurement in

assessing an individual's ability or personality

(Osterlind, 2010). The accuracy of a measurement is

determined by the consistency of measurement

results from various assessments. The more

consistent the measurement results, the better the

reliability. A measuring instrument must have

consistency, so that the result of measuring

instrument from one subject does not have a relatively

different value every time the measuring instrument

is used. Reliability of a measurement result is said to

be reliable if its value is greater than 0.7 (Coaley,

2010).

2 METHODS

2.1 Participants

Two hundred and twenty students of the Faculty of

Psychology Universitas Sumatera Utara (Women =

181, Men = 39) were selected nonrandomly to be

involved in this study. Age ranging from 18 to 23

(average = 19.90, SD = 1.00)

2.2 Instrument

2.2.1 Hamilton Anxiety Rating Scale

(HAM-A)

HAM-A is an instrument developed to measure

anxiety through the severity of anxiety symptoms.

HAM-A is often used in clinical and research

environments. Consisting of 14 items with response

type of Likert Scale which has 5 alternative answers

from no symptoms = 0 to very heavy = 4, with the

total score indicating the severity of anxiety

experienced. Total range of score 0-56 with score

category <17 indicates mild severity, 18-24 indicates

moderate severity and 25-30 indicates severe severity

to very severe. Each item is defined using a series of

symptoms, and measures both psychic anxiety

(mental distress and psychological distress) and

physical anxiety (physical complaints related to

anxiety). This scale has been translated into:

Cantonese, French, and Spayol. This IVR version of

the measuring tool is available at Healthcare

Technology Systems.

2.2.2 Big Five Inventory

The Neurotic subscale is one of five subscales in the

Big Five Inventory (BFI). BFI is based on personality

theory of Big Five model to know the profile of

individual personality. BFI Indonesian version was

developed by Rahmawati and Maryanti (2013). The

Neurotic subscale is used to identify individual

susceptibility to psychological distress: it is easy to

experience sadness, excessive fear and anxiety, has

excessive impulse and has a maladaptive or

inappropriate coping response. This subscale consists

of 8 item using Likert response format with

alternative answer ranging from disagree = 1 to

strongly agree = 5. Total score indicates the level of

vulnerability of individual to distress. The higher the

score is interpreted that the individual has a high

susceptibility to the distress, and the lower the total

score indicates an increasingly stable individual

emotion. The reliability of neurotic subscale in

Indonesian version is 0.762 (Rahmawati, et al, 2016).

In this study, the reliability of the results measured by

Alpha Cronbach of subscale has a good criterion

which is 0.86.

2.3 Procedure

The adaptation process begins with a review of the

coexistence of anxious construct in HAM-A

(Hamilton, 1959) whether it can be used in Indonesia

or not. Phase of language adaptation is done by using

back translation method. HAM-A English version is

translated into Indonesian (translated I), then

translation result I translated back into English

(translation II). The results of translation II are

compared with the original version and analyzed to

ensure that the two forms are equivalent. At this phase

of adapting the language it was also involving 2

people who have expertise in the field of Psychology

to review the content of each item in measuring

anxiety.

ICOSTEERR 2018 - International Conference of Science, Technology, Engineering, Environmental and Ramiﬁcation Researches

1546

The psychometric properties testing phase of

HAM-A uses trial data to find out whether the results

of the HAM-A measurement can reliably measure

anxiety on the respondents. Validity based on internal

structure evidence is done by confirmatory factor

analysis. Estimation of the correlation of HAM-A

was done with the neurotic personality dimension to

prove convergent validity. Testing the mean

differences in the two groups as well as the estimation

of the item difference power to ensure that HAM-A

are truly capable of distinguishing people who

experience and do not experience anxiety. The

accuracy of the measurement results of HAM-A is

known by estimating the Alpha Cronbach coefficient.

3 RESULTS

3.1 Source of Validity Proof based on

Internal Structure

The result of confirmatory factor analysis showed that

62.5% (5 of 8) model matching indexes used showed

good fit with CFI Index = 0.93, IFI = 0.93, GFI = 0.9,

NNFI = 0.91, RMSEA = 0.078, AGFI = 0.86, RFI =

0.86, NFI = 0.89. The 14th factor load is above the

good limit of 0.53 to 0.78 and all the t values are

above the limit of 1.96, moving from 8.13 to 13.57

and all positive errors. The model matching index is

presented in Table 1 and the factor load values for

each item are presented in Table 2. The reliability of

the measurement model is very good (Construct

reliability = 0.99 and Variance extracted = 0.87).

Table 1: Model Matching Index.

No Size Goodness

of Fit

Limit Explanation

1 NFI 0.89 0.80 ≤

NFI ≤

0.90

Marginal

Fit

2 NNFI 0.91 NNFI ≥

0.90

Good Fit

3 RFI 0.86 0.80 ≤

RFI ≤

0.90

Marginal

Fit

4 CFI 0.93 CFI ≥

0.90

Good Fit

5 IFI 0.93 IFI ≥

0.90

Good Fit

6 GFI 0.90 GFI ≥

0.90

Good Fit

7 AGFI 0.86 0.80 ≤

AGFI ≤

0.90

Marginal

Fit

8 RMSEA 0.078 RMSEA

≤ 0.08

Good Fit

Table 2: Characteristics of items in HAM-A.

Item CFA Test of Mean Difference

Difference

power

Reliability If

Item is

Disposed

Factor Load Value of T SE t p Mean

Difference

HAM-A - - - 34.20 0.0001 25.93 - 0.92*

item1 0.54 8.39 0.07 10.29 0.0001 1.58 0.59 0.92

item2 0.68 11.06 0.07 16.35 0.0001 2.18 0.70 0.91

item3 0.63 10.16 0.08 13.36 0.0001 2.12 0.64 0.92

item4 0.65 10.42 0.09 11.52 0.0001 2.15 0.60 0.92

item5 0.60 9.41 0.07 11.60 0.0001 1.80 0.59 0.92

item6 0.78 13.44 0.07 18.26 0.0001 2.35 0.74 0.91

item7 0.77 13.18 0.06 13.84 0.0001 1.93 0.72 0.91

item8 0.78 13.57 0.07 13.30 0.0001 1.97 0.73 0.91

item9 0.70 11.55 0.07 12.00 0.0001 1.92 0.69 0.91

item10 0.73 11.97 0.07 13.32 0.0001 1.85 0.67 0.91

item11 0.72 12.02 0.07 12.88 0.0001 1.92 0.68 0.91

item12 0.68 9.84 0.06 9.85 0.0001 1.45 0.56 0.92

item13 0.69 11.38 0.06 11.09 0.0001 1.70 0.65 0.92

item14 0.53 8.13 0.05 7.77 0.0001 1.02 0.51 0.92

*Reliability of HAM-A scale

The Accuracy of Indonesian Version of HAM-A

1547

3.2 Sources of Validity Evidence based

on Relation to Other Variables

Validity based on the relationship of other variables

is evidenced by correlating the score of HAM-A with

the personality dimension subscale of Neuroticism.

The estimation results show that the result of

measurement using HAM-A correlated significantly

with Neurotic dimension. The large correlation

coefficient is presented in table 3.

Table 3: Correlation of HAM-A and Neuroticism

scale score.

Variabel correlation p Explanation

HAM-

A*Neurotis

0.565

0.0001 Significantly

Correlated

3.3 Reliability

Estimation of reliability coefficient scale of HAM-A

with alpha cronbach is good that is 0.92. Items of

HAM-A also has discrimination power in the

excellent category according to Ebel (Azwar, 2013)

in the range 0.51 to 0.74. In addition to knowing the

ability of item and the scale of HAM-A to distinguish

individuals who are actually experiencing anxiety or

not, a mean difference analysis is performed on high

and low score groups. The results show that there is a

significant difference of mean between high and low

group on all items and scale of HAM-A, with all

differences showing the same direction.

Characteristics for each item and in total can be seen

in table 2.

4 DISCUSSION

HAM-A is a scale for measuring anxiety that has been

widely used in the context of research. HAM-A has

been translated into Cantonese, French, and

Spayol.Adaptation and testing of psychometric

characteristics of HAM-A in this study were

undertaken to ensure that the results of measurements

in Indonesia can be interpreted according to the

original purpose of its manufacture. The validity

analysis to find evidence based on the internal

structure using the data obtained from the field, shows

a good level of compatibility with the existing model.

This is indicated by more than 50% of the model

matching indexes having Good criteria. In addition,

all factor load values obtained in this analysis are

considered satisfactory (> 0.50) Thus it can be said

that all valid items contribute to anxiety

measurement.

Evidence of convergent validity by correlating

HAM-A scores and neurotic dimensions of the Big

Five Inventory shows significant results with

correlations classified as medium. It can be

interpreted that the scores obtained by using HAM-A

really measure participants' anxiety. Neurotics were

chosen as criteria, as many studies show a significant

correlation between anxiety with neurotics, among

which results were found by Johansson & Ölund

(2017) anxiety has been shown to be highly

correlated with neuroticism; neuroticism is

significantly associated with anxiety symptoms

(Daniel et al., 2016), neurotic trait represents a

tendency to experience negative emotions e.g anger,

anxiety, or depression (Kadimpati et al., 2015).

In addition to having evidence of validity based

on internal structure and relationships with other

variables, the reliability of the results of the HAM-A

measurements was found to be excellent. According

to Coaley (2010), the reliability of a measurement

result is said to be reliably if its value is greater than

0.7. When viewed from the item different power

index, all items have an acceptable value. Significant

mean differences between high and low groups in

total scores and all items show that all valid items

contribute to anxiety measurement.

5 CONCLUSIONS

Based on the analysis and review of the results it can

be concluded that the HAM-A which has been

translated into the Indonesian language has an

accurate quality measurement results to measure

anxiety. Thus this scale can be used in studies of

anxiety.

ACKNOWLEDGEMENTS

This paper is basic research scheme funded by

Universitas Sumatera Utara “Penelitian Keunggulan

Akademik (TALENTA) USU T.A USU 2018“.

REFERENCES

Amiri, M., & Ghonsooly, B., 2015. The Relationship

Between English Learning Anxiety And The Students’

Achievement On Examinations. Journal of Language

Teachingand Research, 6, 855 – 865.

ICOSTEERR 2018 - International Conference of Science, Technology, Engineering, Environmental and Ramiﬁcation Researches

1548

Azwar, S., 2013.Penyusunan Skala Psikologi. Pustaka

Pelajar. Yogyakarta.

Burns, D.J., 2004. Anxiety At The Time Of The Final

Exam: Relationships With Expectations And

Performance. Journal of Education for Business, 80,

119-124.

Chapell, M. S., Blanding, Z. B., Silverstein, M. E.,

Takahashi, M., Newman, B., Gubi, A., & McCann, N.,

2005. Test Anxiety And Academic Performance In

Undergraduate And Graduate Students. Journal of

Educational Psychology, 97(2), 268-274.

Coaley, K.., 2010. An Introduction to Psychological

Assessment and Psychometrics. Sage Publication Ltd.

Daniel, J.P.,Salome, V., Peter J.N., and Carla, S., 2016.

Examining unique contributions of three

transdiagnostic vulnerability factors.Personality and

Individual Differences, Volume 94, May 2016, Pages

38-43)

Everson, H. T., Millsap, R. E., & Rodriguez, C. M., 1991.

Isolating Gender Differences In Test Anxiety: A

Confirmatory Factor Analysis Of The Test Anxiety

Inventory. Educational and Psychological

Measurement,51, 243-251.

Hamilton, Max., 1959. The Assessment of Anxiety States by

Rating. The Journal of Medical Psychology, 32, 50-55

Hancock, D. R., 2001. Effects Of Test Anxiety And

Evaluative Threat On Students' Achievement And

Motivation. The Journalof Educational Research, 94,

284-290.

Johansson, M and Ölund, A., 2017. Thinking about

thinking – a study of anxiety, neuroticism and Need for

Cognition.. Tesis. Umeå University.

Kadimpati S, Zale E.L, Hooten M.W, Ditre J.W, Warner

D.O., 2015. Associations between Neuroticism and

Depression in Relation to Catastrophizing and Pain-

Related Anxiety in Chronic Pain Patients. PLoS ONE

10(4): e0126351. doi:10.1371/journal.pone.0126351

Osterlind, S. J., 2010. Modern Measurement: Theory,

Principles, and Application of Mental Appraisal

(Second Edition). United States of America: Pearson

Education, Inc.

Schaefer, A., Matthess, H., Pfitzer, G., & Kohle, K., 2007.

Mental Health And Performance Of Medical Students

With High And Low Test Anxiety. Psychother

Psychosom Med Psychol, 57, 289-297.

Swanson, S., & Howell, C., 1996. Test Anxiety In

Adolescents With Learning Disabilities And Behavior

Disorders. ExceptionalChildren, 62, 389-397.

The Accuracy of Indonesian Version of HAM-A

1549