UNIVERSITAS INDONESIA

PERILAKU OBSESIF-KOMPULSIF: PENGUJIAN RELIABILITAS DAN

VALIDITAS DENGAN VARIABEL LOCUS OF CONTROL, STATUS HUBUNGAN
ROMANTIS, DAN DAFTAR OBSESSIVE-COMPULSIVE INVENTORY (OCI-R)
BARU YANG TELAH DIREVISI

(Obssessive-Compulsive Behaviour: Testing Reliability and Validity with Locus of Control,

Relationship Status, and Obsessive-Compulsive Inventory – Revised)

MAKALAH NON-SEMINAR

NURNABILA KIRONO

1106023713

FAKULTAS PSIKOLOGI

PROGRAM SARJANA

DEPOK

AGUSTUS 2015
   2  

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   3  

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   4  

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   5  

Perilaku Obsesif-Kompulsif: Pengujian Reliabilitas dan Validitas dengan

Variabel Locus of Control, Status Hubungan Romantis, dan daftar
Obsessive-Compulsive Inventory (OCI-R) baru yang telah direvisi.

Nurnabila Kirono1,2, Mark Horswill1, Chiara Santomauro1, Dianti Kusumawardhani2

1.
S1 Kelas Internasional Psikologi, Fakultas Psikologi, Universitas Indonesia, Kota Depok, Jawa Barat,
16424, Indonesia
2.
Undergraduate Program, Bachelor of Arts in Psychology, University of Queensland, St Lucia, QLD
4067, Australia

E-mail: illakirono@outlook.com

Abstrak

Penelitian ini bertujuan untuk mengembangkan skala baru dalam mengukur perilaku obsesif-kompulsif.
Survei diberikan kepada 194 mahasiswa program sarjana S1 dari University of Queensland. Setelah semua data
telah terkumpul dan diselesaikan, skala baru yang telah dikembangkan dari perilaku obsesif-kompulsif pun diuji
reliabilitas, Item Discrimination Index, dan validitas dalam skala hubungannya dengan variabel lain yakni locus
of control, status hubungan romantis, dan daftar Obsessive-Compulsive Inventory (OCI-R) baru yang telah
direvisi. Skala obsesif-kompulsif baru yang telah dikembangkan ternyata memiliki reliabilitas yang tinggi di
antara butir-butir lainnya. Item Discrimination Index menunjukkan bahwa tiap butir skala bisa dibedakan dengan
baik, dengan hanya membutuhkan tiga butir untuk diubah penulisan katanya (M = 32.40, SD = 8.17). Ditemukan
hubungan antara butir skala obsesif-kompulsif dengan locus of control (r(189) = .18, p = .013) dan variabel OCI-
R yang bertugas untuk memvalidasi (r(189) = .37, p < .001), tetapi tidak ditemukan hubungan antara obsesif-
kompulsif dengan status hubungan romatis (r(189) = .08, ns.). Skala butir yang baru dikembangkan pada
penelitian ini terbukti dapat berlaku untuk setting non-klinis serta memerlukan penelitian lebih lanjut untuk
memvalidasi penggunaannya dengan variabel lain yang ada.

Obssessive-Compulsive Behaviour: Testing Reliability and Validity with Locus of

Control, Relationship Status, and Obsessive-Compulsive Inventory – Revised

Abstract

The present study looked to develop a new scale in measuring obsessive-compulsive behaviour.
Surveys were administered to 194 undergraduate university students of The University of Queensland to
complete. After all the data have been gathered, the newly developed scale of OCB was tested on its reliability,
Item Discrimination Index, and validity in conjunction with other variables of locus of control, relationship
status, and OCI-R. The newly developed OCB scale proved to have high reliability among its items and good
discriminating index with only three items that need rewording (M = 32.40, SD = 8.17). Items demonstrated
relationship with locus of control (r(189) = .18, p = .013) and OCI-R validating variables (r(189) = .37, p <

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   6  

.001), but none was found with relationship status (r(189) = .08, ns.). This newly developed scale proved to be
applicable to non-clinical setting and need further researches to validate its use with other existing variables.

Keywords: obsessive-compulsive disorder, locus of control, relationships, reliability, validity.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   7  

Introduction

One of the most widely known classifications of psychopathology that is marked by

obsessions or compulsions that occur persistently, intrusively, and become distressful to the
sufferer is broadly recognized by the name obsessive-compulsive disorder (OCD) (Sica,
Novara, & Sanavio, 2002). Although deemed as rare once, such behaviour has become a
pressing issue in our modern society and has been acknowledged as one of the most popular
disorder that people suffer nowadays with universality ranging up until 2 – 3% in Western
countries (Samuels & Nestadt, 1997; Taylor, 1995). Thus, it is essential to develop construct
to measure symptoms that extend beyond clinical populations not only for the good of others
suffering from this disorder but to also deepen our understanding of this disorder (Hajcak,
Huppert, Simons, & Foa, 2004).

Theoretical Framework

Numerous investigators are constantly on the search for new treatments to improve
efficacy in treating mental illnesses specifically obsessive-compulsive disorder (OCD), but at
the end of the day these attempts require reliable and valid instruments to correctly assess
targeted traits (Taylor, 1995). Rassin and Muris (2006) claimed that obsessions are popular in
non-clinical situations, however slight over-interpretations by numerous studies regarding
what OCD actually is have occurred. Taylor (1995) have also argued that instead of focusing
on measures of obsessive-compulsive personality traits, studies should focus on measures of
the symptoms if they wish to look at non-clinical sample as well as to avoid over-
interpretations. Therefore, to avoid over-interpretation of targeted behaviour, this study aimed
to assess scales that would correctly evaluate symptoms of obsessive-compulsive behaviour
and is administrable to a non-clinical population. Aside from that, the reliability, validity, as
well as quality of each item in the newly developed scale would be assessed to know which
items are good at measuring the targeted trait and which are not.
Previous scale used Leyton Obsessional Inventory (LOI) to measure perfectionist
housewives and to measure clinical obsessive-compulsive episodes (Cooper, 1970; Cooper &
McNeil, 1968). It consisted of 69 items which comprised of 46 items that measure symptoms
of obsessive-compulsive and 23 items that describe obsessive-compulsive trait (Taylor,
1995). LOI proved to have good psychometric properties with high internal consistencies

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   8  

(Cronbach’s α = .75 to .90), good test-retest reliabilities over 14-day interval, correlation of r
= .87 for the symptom score, and correlation of r = .91 for trait score, as well as establishing
good discriminant validity (Cooper, 1970; Taylor, 1995).
However, LOI has some limitations which is it was originally developed for
housewives as opposed to real patients which may hint that it could not be applicable for
clinical OCD patients when directly compared to its non-clinical counterparts (Foa, Kozak,
Coles, Amir, Salkovskis, 1998). Another problem that was found was that the procedure of
LOI was considered inefficient and unable to detect symptom reductions (Foa et al., 1998).
For that reason, this study wanted to develop a new measure that assess obsessive-compulsive
behaviour that was validated with other measures of personality as well as non-personality
trait and not to mention to have a high validation measure when compared to previously
existing scale.
The present study sought to establish a new scale to measure obsessive-compulsive
behaviour. To assess the reliability of the item, a calculation of internal consistency would be
performed to get an idea of the overall consistency of this scale. To test the newly developed
scale’s degree of validity, this new scale would be evaluated by assessing its correlation with
three other validating scales which are locus of control that assesses participants’ level of
religiousity, participants’ romantic relationship status, and a previously existing scale of
Obsessive-Compulsive Inventory – Revised.
Locus of Control. Several studies have measured the link between religiousity as
individuals’ locus of control and obsessive-compulsive behaviour. Rachman (1997) claimed
that people who have been taught or have learned that their value-packed thoughts are thought
as significant to them and their surroundings would be more susceptible to having obsessions,
such as those who hold some type of religious beliefs and practices. From this it has been
inferred that measures of obsessive-compulsive symptoms to be positively correlated with
religiousity in patients who suffer from OCD which suggested that there are some association
between religiousity and OCD symptoms (Sica, Novara, & Sanavio, 2002).
This can be due to some religious teachings that mostly fall into the domain that have
been deemed to be related and relevant to OCD symptoms, for example beliefs that teach to
overemphasize on the need to control one’s behaviour and thought (Rachman, 1970). Hence,
this study hypothesizes that higher level of OCB symptoms would be positively correlated
with higher claim of participants’ religiousity level, such that if this test proves to be valid,
those who score higher on the newly developed OCB scale would also report higher level of
religiousity.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   9  

Romantic Relationship Status. Few studies have examined the link between
obsessive-compulsive disorder with individuals’ romantic relationships (Abbey, Clopton, &
Humphreys, 2007). Difficulties in interpersonal relationships and impairment in social
functioning often is linked with obsessive-compulsive disorder (OCD) and even symptoms
that are considered subclinical in OCD (Angst et al., 2004; Bystritsky, 2001). Particularly,
severity of OCD symptom may be related to satisfaction in relationship such that married
individuals who suffer from OCD often display significant amount of marital distress as well
is the case as in those with OCD often are not married (Abbey, Clopton, & Humphreys, 2007;
Riggs, Hiss, & Foa, 1992). Therefore, this study predicts that participants’ relationship status
(single, dating, in a long term relationship, married or defacto) would be negatively correlated
with obsessive-compulsive behaviour (OCB). Such that if proven valid, those who score
highly in the OCB would report lower commitment level in relationship status scale.
Obsessive-Compulsive Inventory – Revised (OCI – R). Lastly, when compared to
the original Obsessive-Compulsive Inventory (OCI), the OCI-R is significantly shorter
(Hajcak, Huppert, Simons, & Foa, 2004). The OCI-R scale has removed items that seemed to
be unnecessary and has subscales that comprised of three summed items which make them
easier to compare (Hajcak, Huppert, Simons, & Foa, 2004). This revised scale proved to
obtain very good properties when assessed on mixed sample patients with OCD, other anxiety
disorders, as well as to non-patients (Hajcak, Huppert, Simons, & Foa, 2004).
Even though more studies need to be conducted, but this sort of scale and research
have proven to be an important basis for theoretical understanding of OCD although
administered on non-clinical samples (Hajcak, Huppert, Simons, & Foa, 2004). For that
reason, this study predicts that there would be a positively correlated relationship between the
newly developed scale of OCB with OCI-R such that if measures of the scales are valid
people who score higher on OCB would also demonstrate higher score on OCI-R.

Methods

A total of 194 people that comprised of 144 female and 50 male university students
from The University of Queensland who enrolled in Measurement in Psychology course
participated in this study. The mean age of the participants was 22.14 (SD = 3.95). Out of 194

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   10  

people that participated, 191 completed questionnaires were returned.

Demographic information. Participants were required to state their age upon

completing the questionnaire. Participants were also required to indicate their sex where
female was coded as 1 and male was coded as 2 in the data. Since they are only single-item
scales, internal consistency was not measured for both age and gender.

Obsessive Compulsive Behaviour (OCB). A newly developed scale consisted of 12

items was used to measure participants’ level of obsessive-compulsive behaviour. These
scales were generated by students during their tutorial classes where they were divided into
groups of 3-4 people. Each group was instructed to assemble 3-4 statements associated with
OCB-related symptoms in everyday life which revolved around the specific traits of phone
checking behaviours. These collected statements were then collated into the final tutorial
group questionnaire chosen by the tutors as the pilot data. After that they were to discuss and
chose twelve items that deemed as suitable to be utilized later as the statements for the newly
developed pilot data questionnaires.
Nine items were measuring OCB right away using statements like, “I often feel
anxious if I am unable to check my phone for notifications,” or “I am often distracted by
thoughts about the social applications on my phone.” While three out of twelve items were
scored reversely, for example using the statement, “if I accidentally leave my phone at home
it does not bother me.” The response scale ranged from 1 = strongly disagree to 5 = strongly
agree for all of the items, with higher score indicating a higher measure of OCB. This test
yielded high internal consistency (Cronbach’s α = .85).

Locus of Control. This measure utilized a single item self-report scale that required
participants to rate themselves on how religious they are. Participants responded with scale
that ranged from 1 = not at all to 5 = very much so. Higher score dignified a higher level of
religiousity as self-reported by participants. Internal consistency was not measured since it
was a single-item scale.

Romantic Relationship. As part of another validating scale, a demographical variable

of participants’ romantic status were required to be completed. Participants were instructed to
indicate their romantic status that is coded as 1 for single, 2 for dating, 3 for long term
relationship, and 4 for married or de-facto in the data. The internal consistency was not
measured for this single-item scale.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   11  

Obsessive Compulsive Inventory-Revised (OCI-R). Twelve items out of the

original eighteen items derived from Foa et al. (2002) were used as a validating measure to
evaluate OCB. Using five points Likert-type scales that scored from 1 = not at all to 5 =
extremely, higher scores indicated a higher level of OCB. All of the items directly measured
OCB with sample statement such as, “I collect things I don’t need,” and reversed items were
not utilized here. These items demonstrated high internal consistency (Cronbach’s α = .87).
Participants completed questionnaires during their free time online and all the data
that was administered was collated altogether. Participants’ contribution in this study was
strictly voluntary and they were free to withdraw from participation at any time with no
penalty. All information retained in this study was also promised anonymity. This study was
administered and conducted as a major compulsory assignment among undergraduate
students who were taking Measurement in Psychology course in order to pass this course in
pursuit of their undergraduate degree in Psychology/Arts major.
Table 1 demonstrated that Obsessive-Compulsive Behaviour (OCB) scored the highest
with mean of 32.40 (SD = 8.17). OCB also has the biggest range of scores which varies from
14 to 53. Since all of the data from the scales were normally distributed, transformations for
skewness were not performed. Pearson correlation analysis was performed in order to see the
relationship between the newly developed Obsessive-Compulsive Behaviour scale and three
other validating scales which are Locus of Control, romantic status, and a previously existing
scale to measure OCB that is Obsessive-Compulsive Inventory-Revised (OCI-R).
Furthermore, the newly developed scale was assessed with the three other validating scales to
ensure high reliability first with regards to its Cronbach’s alpha scores in order to establish a
good measure of validity, because a test of reliability is prerequisite to validity (Webb,
Shavelson, & Haertel, 2006).

Results

Table 1. Descriptive Statistics for Obsessive-Compulsive Behaviour (OCB), Locus of Control, Romantic
Status, and Obsessive-Compulsive Items-Revised (OCI-R)

Variable M SD Minimum Maximum Skew

OCB 32.40 8.17 14.00 53.00 0.20
Locus of Control 2.22 1.20 1.00 5.00 0.79
Romantic Status 1.89 1.00 1 4 0.56
OCI-R 24.29 8.10 12.00 49.00 0.86

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   12  

There was a positively significant correlation between OCB and Locus of Control
which measured participants’ self-reported level of religiousity (r(189) = .18, p = .013), such
that as participant’s level of religiousity increases it is linked with higher level of OCB. No
significant correlation was found between OCB and romantic relationship status of
participants (r(189) = .08, ns.). Furthermore, OCB and OCI-R was found to be significantly
positively correlated (r(189) = .37, p < .001), such that higher scores of OCB is linked with
higher score in OCI-R scales. Following that, Item Discrimination Indices were calculated to
analyze the quality of each items in the newly developed scale of OCB. The goal is to know
to what extent does one item from the scale would be able to identify between participant who
score low and high on average within this test.
The steps to calculate the Item Discrimination Index is by first identifying the bottom
and the top 33% of scorers which will be coded as top scoring group (nU) and low scoring
group (nL). Then for each item in the scale, divide the number of high scorers that obtained
the item correctly (U) with the total number of high scorers (nU). Next, divide the number of
low scorers that obtained the item correctly (L) with the total number of low scorers (nL), and
then subtract the first high scorer result with this result, thus Item Disctimination Index score
(d) would be obtained. As can be seen from Table 2, item 4 from the OCB scale has the
highest range of scores out of the rest and the Item Discrimination Indices score indicate
scores that are closest to one (d = .67). While the lowest range of score fall to item 1 (d = .20)
(See Table 2).

Table 2. Item Discrimination Indices for Obsessive-Compulsive Behaviour Scale

Variable U L d
TuteOCB_1 45 46 0.20
TuteOCB_2 52 38 0.44
TuteOCB_3 21 0 0.38
TuteOCB_4 42 7 0.67
TuteOCB_5 19 3 0.31
TuteOCB_6 40 6 0.65
TuteOCB_7 20 0 0.36
TuteOCB_8 49 23 0.58
TuteOCB_9 21 1 0.37
TuteOCB_10 35 0 0.64
TuteOCB_11 29 1 0.51
TuteOCB_12 34 3 0.58

Note. nU = 55 and nL = 75 for all items.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   13  

Discussion

This finding is consistent with previous study that stated some association between
level of religiousity and OCD symptoms may exist such that the more religious the individual
is the more susceptible they will be to having obsessions, such as those who hold some type
of strict religious beliefs and practices (Rahman, 1997; Sica, Novara, & Sanavio, 2002). It can
be inferred that this newly developed scale is a good and valid measure of OCB when
compared with religiousity.
Second, this study revealed that there is no relationship between OCB and relationship
status. This is contradictory to what previous findings have suggested whereby have been
stated that those with OCD often are not married (Abbey, Clopton, & Humphreys, 2007). It
also is contrary to what have been said previously about marital satisfaction as well as the link
between difficulties in interpersonal relationships and impairment in social functioning that
was said to be linked with symptoms that are considered apparent in non-clinical sufferers in
OCD (Angst et al., 2004; Bystritsky, 2001; Riggs, Hiss, & Foa, 1992). Thus suggesting that
perhaps more should be looked into the scale and its items before assessing it as a validating
variable with romantic relationship status.
Third, it was revealed that the newly developed OCB scales proved to be valid when
measured with the old OCB revised inventory (OCI-R). This suggests that since the old OCI-
R scale proved to have very good properties when assessed on mixed samples of patients with
OCD, other anxiety disorders, as well as to non-patients, the newly developed OCB scale also
proved to be doing the same thing especially to the non-clinical population (Hajcak, Huppert,
Simons, & Foa, 2004). Thus this may suggest that the newly developed measure of OCB
would be applicable and valid to use among non-clinical population to measure subclinical
symptoms of OCB.
The internal consistency was also assessed and it revealed that all items in the newly
developed OCB scale demonstrated high reliability. Next, Item Discrimination Index revealed
that from all of the OCB scale, item 4 has the highest range of scores out of the rest and
indicate that most high scorers got this items right and low scorers got this item wrong. The
lowest range of score fall to item 1, 5 and 6 which means these items would possibly need to
be removed or reworded because they indicate their inability to differ between high and low
scorer by looking at previous researches to see what specific variables could be used as
statement items to measure OCB.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   14  

Conclusion

This study aimed to develop a newly developed and improved construct to measure
obsessive-complusive behaviour (OCB). In order to assess a good measure of validity, the
newly developed scale was assessed to ensure high reliability first with regards to its
Cronbach’s alpha scores with the three other variables which are locus of control, relationship
status, and a previously existing revised inventory of obsessive-complusive (OCI-R). OCB
was found to be associated with locus of control that measure participants’ self-reported level
of religiousity. Such that, people who score higher on their newly developed OCB scale
would also demonstrate higher level of religiousness.

Suggestion

The key strength of this study was the internal consistency between variables was high
indicating each items were correct at measuring targeted trait. As well as number of items that
is considerably few which makes this measure easy to perform for most people. Which could
be another strength of this study, that is its ability to be used in non-clinical setting. One of
the limitations of this study would be that participants were given too many items to answer
in one session, which could affect the way participants answered the survey due to possible
boredom or tiredness. The generalization of the result would also be restricted due to its
administration that was distributed only to a sample of university students located in Australia
undergoing certain course in their undergraduate study.
For future research, instead of using self assessing measure to assess religiousity,
researchers perhaps could utilize items from previous studies or other measuring scale that are
more objective and are not biased towards what participants’ own interpretation of their level
of religiousity. And perhaps the newly developed OCB scale could be narrowed down with
the help of more findings and trial in future research, as well as could be narrowed down and
made flexible for its usage in various samples. Lastly, the present study sought to establish a
new scale to measure obsessive-compulsive behaviour. The scale demonstrated high
reliability and validity in conjunction with measure of locus of control as well as previously
existing scale but proved to display no relationship with relationship status.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   15  

References List

Abbey, R. D., Clopton, J. R., & Humphreys, J. D. (2007). Obsessive-compulsive

disorder and romantic functioning. Journal of Clinical Psychology, 63(12),
1181-1192.
Angst, J., Gamma, A., Endrass, J., Goodwin, R., Ajdacic, V., Eich, D., & Rössler, W.
(2004). Obsessive-compulsive severity spectrum in the community: Prevalence,
comorbidity, and course. European Archives of Psychiatry and Clinical Neuroscience,
254(3), 156-164.
Bystritsky, A., Liberman, R. P., Hwang, S., Wallace, C. J., Vapnik, T., Maindment,
K., & Saxena, S. (2001). Social functioning and quality of life comparisons between
obsessive-compulsive and schizophrenic disorders. Depression Anxiety, 14(4), 214-
218.
Cooper, J. (1970). The Leyton obsessional inventory. Psychological Medicine, 1, 48-

64.
Cooper, J., & McNeil, J. (1968). A study of house-proud housewives and their
interaction with their children. Journal of Child Psychology and psychiatry, 9, 173-
188.
Foa, E. B., Kozak, M. J., Coles, M. E., Amir, N., & Salkovskis, P. M. (1998). The
validation of a new obsessive-compulsive disorder scale: The obsessive-
compulsive inventory, Psychological Assessment, 10(3), 206-214.
Foa, E. B., Huppert, J. D., Leiberg, S., Langner, R., Kichic, R., Salkovskis, P. M., &
Hajcak, G. (2002). The obsessive-compulsive inventory: Development and validation
of a short version. Psychological Assessment, 14(4), 485-496.
Hajcak, G., Huppert, J. D., Simons, R. F., & Foa, E. B. (2004). Psychometric
properties of the OCI-R in a college sample. Behaviour Research and Therapy, 42,
115-123.
Rassin, E., & Muris, P. (2006). Abnormal and normal obsessions: A reconsideration.
Behaviour and Research Therapy, 45, 1065-1070.
Riggs, D. S., Hiss, H., & Foa, E. B. (1992). Marital distress and the treatment of
obsessice compulsive disorder. Behavioural Therapy, 23, 585-597.
Samuels, J., & Nestadt, G. (1997). Epidemiology and genetics of obsessive-
compulsive disorder. International Review of Psychiatry, 9, 61-71.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015

   16  

Sica, C., Novara, C., & Sanavio, E. (2002). Religiousness and obsessive-compulsive
cognitions and symptoms in an Italian population. Behaviour Research and Therapy,
40, 813-823.
Taylor, S. (1995). Assessment of obsessions and compulsions: Reliability, validity,
and sensitivity to treatment effects. Clinical Psychology Review, 15(4), 261-296.
Webb, N. M., Shavelson, R. J., Haertel, E. H. (2006). Four reliability coefficients and
generalizability theory. Handbook of Statistics, 26, 81-124.

Perilaku obsesif…, Nurnabila Kirono, Fpsi UI, 2015