Development of The Addiction Profile Index Internet Addiction Form (APIINT): validity and reliability
Kultegin Ogel, Figen Karadag, Defne Satgan, Ceren Koc
Article No: 5   Article Type :  Research
Objective: To adapt the Addiction Profile Index (API) for internet use and addiction in order to develop an internet addiction questionnaire better suited to the definition of addiction to be used for clinical purposes.

Method: The questions of the API were adapted to measure the use of the internet. After pilot application and feedback, a new scale of 18 items was developed. Included in the study were 103 randomly selected first-year students of a private high school and the entire first year cohort of a medical faculty (n=51). The Internet Addiction Scale developed by Nichols and adapted to Turkish by Canan was used for the validation of the questionnaire. The questionnaire was reapplied to the subjects fifteen days later.

Results: Cronbach’s ? coefficient for the total APIINT was 0.88, and retest correlation was 0.85. Three factors were obtained using explanatory factor analysis that represented 57.03% of the total variance. A correlation coefficient of 0.81 was found between APIINT and Internet Addiction Scale. The area under the ROC curve was 0.97.With a total APIINT cut-off score of 2, both the scale’s sensitivity and specificity were 0.90, respectively. A screening form consisting of 2 questions with a cut-off score 3.5 had a sensitivity of 0.72 and its specificity was 0.83.

Conclusion: The results show that the APIINT is a valid and reliable questionnaire that can be used for high school and university students.
Keywords : Addiction, internet, scale
Dusunen Adam : The Journal of Psychiatry and Neurological Sciences : 2015;28:337-343
Full Text:


Internet addiction is not yet a fully recognized diagnostic category. In studies and development of scales, a variety of terms is being used, such as computer addiction, virtual addiction, and technology or game addiction. It has been shown that internet addiction is a common problem that can negatively affect people’s academic, social, economic, and family lives (1,2). It has been pointed out that factors like the ready availability and low cost of the internet and the opportunity for individuals to create an identity outside their own while on the internet may contribute to the fast increase of the problem’s dimension (3).

While problems related to internet use are found to resemble impulse control disorders or gambling addiction, it has been proposed that they are psychopathologically distinct from these categories (4,5) and show more similarity with alcohol and substance addition (6,7). In the latest published classification, it has been assessed as “internet game playing addiction” (8). Generally speaking, scales concerned with internet addiction are to a large extent coinciding with scales used to diagnose addiction (9,10).

Several scales have been developed to deal with internet addiction (11). In Turkey, too, some scales to assess internet addiction have been developed and others have been translated into Turkish and tested for their validity and reliability (12-14).

The Addiction Profile Index (API) is a scale developed to assess the severity of addiction and its various dimensions (15). This scale has been used in a number of clinical studies in Turkey. Aim of our study was to adapt API to the assessment of internet use and to develop an internet form for the instrument.


Internet Addiction Scale: Developed by Nichols (16,17) and adapted to Turkish by Canan (18), is a scale consisting of 27 items. Its internal consistency is fairly high (Cronbach’s α=0.94) and the one-week test-retest reliability is also quite good (Spearman-Brown correlation coefficient: 0.98; p<0.001). Validity and reliability of the scale have been tested for high schools as well as for university students (18). Cut-off point is a score of 0.81.

Addiction Profile Index (API): A self-reporting scale consisting of 37 questions and 5 subscales measuring characteristics of substance use, dependency diagnosis, the effects of substance use on the user, craving, and motivation to quit using substances (15). In the reliability analysis, Cronbach’s α for the entire scale was 0.89 and for the subscales, it ranged between 0.63 and 0.86. The item-total score correlation coefficient was between 0.42 and 0.89. For the whole scale, the Spearman-Brown coefficient for the split half-test correlation was 0.83. In the explanatory factor analysis, 4 factors representing 52.39% of the variance were obtained. The API Craving subscale was correlated with the Penn Alcohol Craving Scale (PACS) and motivation subscale with the Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES). The API total score was correlated with the mean score in the Michigan Alcohol Screening Test (MAST) and the composite score in the Addiction Severity Index (ASI).


One hundred and three randomly selected first-year students at a private high school and the entire first year cohort of a university medical school (n=53) were included in the study. All students enrolled in the study (n=154) completed the scale. There were no forms that had to be excluded for having been filled in incompletely or wrongly. As this was a standardization study, no distinction was made between the participants. For the study, approval was received from the ethics committee of Acibadem University. The forms were completed by the students anonymously and mixed during collection. Average age of the high school students was 17.7±0.6 years, mean age of the university students was 20.0±0.6 years.


After completing the forms, 108 students filled in the forms once again 15 days later.

Development of the Scale

Since internet addiction was a category for which no scales existed, it has been suggested to use addiction scales as a foundation (10). Thus, in order to develop an internet addiction scale, the Addiction Profile Index (API) was used, for which validity and reliability studies in Turkey existed. Questions for the scale were prepared in accordance with the dimensions of API. However, in order to assess internet issues, the questions were asked in a different structure. Questions not suitable for an investigation of internet addiction (e.g., “Are there other substances you are using?”) were removed from the forms. Answer choices were presented, like in the API, on a 5-point Likert-type scale. The question form thus developed was administered by teachers to 10 persons who were known as intensive internet game players. Each question was reviewed individually, and the testers’ views and feedback were collected.

According to the feedback received, the structure of the questions was modified and questions not understood by the testers were removed from the form. The completed questions were sent to 5 experts working in the field of addiction and their views were received. The questionnaires were finalized in accordance with the feedback. Thus, face validity was obtained. Cronbach’s α coefficient for the scale consisting of 21 questions and 5 dimensions was found to be 0.86. It was agreed to remove 3 items with a reliability coefficient below 0.4 from the scale (“You wanted to use the internet, but when you could not use it, did any problems emerge?”, “Has internet use affected your bodily health negatively?”, and “Were you upset when you could not get on the internet when you had planned to?”) The resulting scale consisting of 18 questions and 5 dimensions was named Addiction Profile Index Internet Form (APIINT).

Statistical Analysis

For the reliability analysis, Cronbach’s α coefficient was calculated, and the factor structure of the scale was evaluated using explanatory factor analysis and varimax rotation. The scale’s criterion-related validity analysis was performed using Pearson correlation analysis. All statistical calculations in this study were made using the SPSS 17.0 package.



Cronbach’s α coefficient for the entire scale was 0.88 and for the subdimensions it ranged between 0.64 and 0.77. Item-total score correlation coefficients were between 0.44 and 0.68 (Table 1).

The retest correlation for the entire scale was 0.85 (p<0.01). Retest correlations for the subdimensions were all statistically significant (p<0.01) and ranged from 0.64 to 0.82.


Explanatory factor analysis was performed using the principal component method and varimax rotation. In the factor analysis, like in the original scale, internet use characteristics were not included in the analysis. In the explanatory factor analysis, 4 factors with an eigenvalue above 1 were obtained, accounting for 57.4% of the variance (Table 2).

The first factor can be named “impact of internet use on life”, the second factor “diagnostic criteria for addiction”, the third factor “motivation”, and the fourth factor “stopping to attend events”. Considering the fourth factor to contain one of the diagnostic criteria, the average of the 3 questions contained in this factor was used to create the second factor. Thus, 3 factors explaining 57.03% of the total variance were found. Questions relating to “stopping to attend events”, which in the previous analysis was seen as a separate factor, were now included in the factor “impact of internet use on life” (Table 3).

The explanatory factor analysis carried out with the total score from the subdimensions resulted in only one factor with an eigenvalue above 1, accounting for 57.3% of the total variance. Factor loads were 0.88 for diagnostic criteria, 0.84 for impact on life, 0.80 for craving, 0.62 for internet use characteristics, and 0.61 for motivation, respectively.

Scale-Related Validity

The correlation coefficient between APIINT and Internet Addiction Scale was 0.81 (p<0.01). Based on the Internet Addiction Scale, the area under the ROC curve (Figure 1) was 0.97 (p<0.001). With a cut-off point of 2, APIINT sensitivity was 0.90, and sensitivity, too, was 0.90 (positive predictive value [PPV]: 99.1 and negative predictive value [NPV]: 61.5). In the study, 83% of the participants were under cut-off point of 2.

Two questions were selected to be used for a screening form: “duration of internet use” and “internet use can cause problems in a person’s life”. These questions show a high correlation with the total overall APIINT score (0.82, p<0.01). Using the Internet Addiction Scale as a basis, the area under the ROC curve (Figure 2) is 0.88 (p<0.001). With a cut-off point at a score of 3.5, sensitivity is 0.72, specificity 0.83 (PPV: 97.5, NPV: 75.7).

High School and University Students

In a separate analysis for the high school students, Cronbach’s α coefficient was 0.88. In the explanatory factor analysis, four factors with an eigenvalue above 1 were found, accounting for 59.2% of the total variance. Distribution of questions to factors was similar to that in the general group. Cut-off point was 2.

In a separate analysis for the university students, Cronbach’s α coefficient was 0.90. In the explanatory factor analysis, four factors with an eigenvalue above 1 were found, accounting for 71.7% of the total variance. Distribution of questions to factors was similar to that in the general group. Cut-off point was 2.


After factor analysis of the questions, it can be said that distribution to factors was consistent with dimensions identified in earlier studies with other API forms. As a result of factor analysis, questions related to “stopping to attend events” were transferred from the “diagnostic criteria” dimension to “impact on life”. It has been noted that in the original API paper, “stopping to attend events” also remained in between two dimensions (15). While various studies found similar results (20), we believe that in research on groups with high-intensity addiction, attending events may drop.

In the factor analysis, craving fell into the diagnostic criteria dimension. However, we believe that in order to facilitate patient assessment in clinical use, it is beneficial to leave craving as a separate dimension. Thus, we can say that APIINT consists of the dimensions listed below:

•Frequency of internet use

•Diagnostic criteria of addiction (addiction symptoms)

•Impact of internet use on life

•Craving for internet use

•Motivation to reduce internet use

The subdimensions of APIINT resemble those put forward by a number of other clinical studies (21,22). All studies specified that symptoms like impact of internet use on life, frequency of internet use in daily life, development of tolerance, or deprivation could be used to identify the problem. The diagnostic criteria proposed for internet addiction are also similar (9).

In separate evaluations for high school and university students, reliability coefficients reached a good level, and the factor structure in both populations was similar. Thus, we can say that the scale can be used in either group. We do believe, though, that for using the scale with adults, respective standardization studies need to be conducted.

In our view, data obtained from applying APIINT in larger and more socioeconomically diverse populations will contribute to the scale’s validity and reliability. It will be especially useful, after developing diagnostic criteria for internet addiction, to compare scores for groups that are and those that are not addicted. In the current state, we must not forget that scale reflects a diagnostic category whose diagnostic criteria are not entirely clear, and we have to evaluate it accordingly.


1.Young KS. Internet addiction: the emergence of a new clinical disorder. Cyberpsychol Behav 1998; 1:237-244. [CrossRef]

2.Ko CH, Yen JY, Chen CC, Chen SH, Yen CF. Proposed diagnostic criteria of Internet addiction for adolescents. J Nerv Ment Dis 2005; 193:728-733. [CrossRef]

3.Cooper A. Sexuality and the Internet: Surfing into the new millennium. Cyberpsychol Behav 1998;1:181-187. [CrossRef]

4.Dalbudak E, Evren C, Topcu M, Aldemir S, Coskun KS, Bozkurt M, Evren B, Canbal M. Relationship of Internet addiction with impulsivity and severity of psychopathology among Turkish university students. Psychiatry Res 2013; 210:1086-1091. [CrossRef]

5.Tonioni F, Mazza M, Autullo G, Cappelluti R, Catalano V, Marano G, Fiumana V, Moschetti C, Alimonti F, Luciani M, Lai C. Is Internet addiction a psychopathological condition distinct from pathological gambling? Addict Behav 2014; 39:1052-1056. [CrossRef]

6.Griffiths M. Does internet and computer ‘addiction’ exist? Some case study evidence. Cyberpsychol Behav 2000; 3:218-222. [CrossRef]

7.Hollander E, Stein DJ. Clinical Manual of Impulse-control Disorders. Arlington, VA: American Psychiatric Publishing, 2006; 184-189.

8.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 5 (DSM-5). Arlington, VA: APA, 2013.

9.Tao R, Huang X, Wang J, Zhang H, Zhang Y, Li M. Proposed diagnostic criteria for internet addiction. Addiction 2010; 105:556-564.[CrossRef]

10.Van Rooij AJ, Prause N. A critical review of “Internet addiction” criteria with suggestions for the future. J Behav Addict 2014; 3:203-213.[CrossRef]

11.Aboujaoude E. Problematic internet use: an overview. World Psychiatry 2010; 9:85-90. [CrossRef]

12.Sahin C. An analysis of internet addiction levels of individuals according to various variables. TOJET: The Turkish Online Journal of Educational Technology 2011; 10:60-66.

13.Ceyhan E, Ceyhan AA, Gurcan, A. The validity and reliability study of Problematic Internet Use Scale. Educational Sciences: Theory and Practice 2007; 7:387-416. (Turkish)

14.Kayri M, Gunuc S. The profile of internet dependency in Turkey and development of internet addiction scale: study of validity and reliability. Hacettepe University Journal of Education 2010; 39: 220-232. (Turkish)

15.Ogel K, Evren C, Karadag F, Gurol T, Satgan D. The development, validity, and reliability of the Addiction Profile Index (API). Turk Psikiyatri Derg 2012; 23:264-273. (Turkish)

16.Griffiths M. Internet addiction: Does it really exist? In Gackenbach J (editor). Second Ed. New York: Academic Press, 1998, 61-75.

17.Nichols LA, Nicki R. Development of a psychometrically sound Internet addiction scale: a preliminary step. Psychol Addict Behav 2004; 18:381-384. [CrossRef]

18.Canan F, Ataoglu A, Nichols LA, Yildirim T, Ozturk O. Evaluation of psychometric properties of the internet addiction scale in a sample of Turkish high school students. Cyberpsychol Behav Soc Netw 2010; 13:317-320. [CrossRef]

19.Canan F. Relationship between internet addiction, disassociative symptoms and sociodemographic features in university students. Thesis, Duzce University, Faculty of Medicine, Department of Psychiatry, Duzce 2010. (Turkish)

20.Senormanci O, Konkan R, Guclu O, Senormanci G. Evaluation of coping strategies of male patients, being treated in internet addiction outpatient clinic in Turkey. Journal of Mood Disorders 2014; 4:14-19. [CrossRef]

21.Shapira NA, Lessig MC, Goldsmith TD, Szabo ST, Lazoritz M, Gold MS, Stein DJ. Problematic internet use: proposed classification and diagnostic criteria. Depress Anxiety 2003; 17:207-216. [CrossRef]

22.Aboujaoude E, Koran LM, Gamel N, Large MD, Serpe RT. Potential markers for problematic Internet use: a telephone survey of 2,513 adults. CNS Spectr 2006; 11:750-755.

11th National Alcohol and Substance Abuse Congress
Creative Commons Lisansı

Dusunen Adam: The Journal of Psychiatry and Neurological Sciences is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Düşünen Adam - Psikiyatri ve Nörolojik Bilimler Dergisi
Bakırköy Prof. Dr. Mazhar Osman Ruh Sağlığı ve Sinir Hastalıkları Eğitim ve Araştırma Hastanesi
Yerküre Tanıtım ve Yayıncılık Hizmetleri A.Ş.