The Assessment of Problematic Internet Pornography Use: A Comparison of Three Scales with Mixed Methods

The primary aim of this study was to compare different screening tools for problematic internet pornography use (IPU) and identify the most accurate measure. The reliability and validity of three scales, namely, the Problematic Pornography Consumption Scale (PPCS), Problematic Pornography Use Scale (PPUS), and Short Internet Addiction Test Adapted to Online Sexual Activities(s-IAT-sex), were examined using three homogeneous groups, respectively. A total of 972 adults (mean age = 24.8) from 28 provinces/regions in China participated in the quantitative part (QUAN). The Brief Pornography Screener served as the reference standard. The PPCS demonstrated stronger reliability and validity, including criterion validity, as well as greater sensitivity and acceptable specificity; therefore, it was considered to be the more accurate screening instrument. In the qualitative part (QUAL), we interviewed 22 volunteers and 11 therapists (who had worked with individuals with problematic IPU) to examine their perspectives on the core features of problematic IPU and dimensions of the PPCS. Almost all the interviewees endorsed the structure of the PPCS. These findings encourage the use of the PPCS in future research studies and underscore its screening applications because of its ability to classify IPU as problematic or nonproblematic.


Introduction
Internet pornography use (IPU) is a sexual behavior [1], corresponding to the use of internet to engage in various gratifying sexual activities also known as online pornography use or cybersex [2][3][4]. It comprises a variety of online sexual activities (OSAs), including watching pornography, online pornography exchange, engaging in sex chats, using sex webcams, searching for sexual partners, or engaging in sexual role playing, among which stands the watching pornography, which is the most popular activity [5]. According to the past findings, engaging in IPU sometimes derives various negative consequences, such as financial, legal, occupational, and relationship trouble or personal problems [6]. Feelings of loss of control and persistent use despite these adverse outcomes constitute compulsive cybersex or problematic IPU. To date, no consensus exists regarding the conceptualization and diagnosis of problematic IPU. For instance, numerous terms have been used to describe the phenomenon (e.g., internet sex addiction [7,8], problematic online sexual activities [9], cybersex addiction [10], and problematic internet pornography use [6]). Although these concepts are slightly different, they all comprise three crucial components: the medium (the internet), the content (sexual behavior), and the problematic use (the compulsive behavior). Regardless of the debate, it is now acknowledged that excessive involvement in IPU or cybersex may become dysfunctional and associated the Problematic Pornography Consumption Scale (PPCS); this scale is founded upon the theoretical framework of Griffiths's components model of addiction [22]. All three scales include strong internal consistency and a valid factorial structure, which has been supported by the results of confirmatory factor analysis (CFA) [9,14,15,19]. Nevertheless, it is difficult to compare the findings of studies that have used these scales because they entail different factor structures. Therefore, it is necessary to select reliable indicators and methods, and identify the most accurate instrument.
In order to effectively compare different scales, a unifying and reliable standard should first be established. The Brief Pornography Screener (BPS), which is a screening tool that measures loss of self-control, overuse of problematic pornography use, may be useful in identifying individuals who are at risk for problematic pornography use or can serve as a proxy measure [23]. Kraus et al., who developed the BPS, have proposed that the diagnostic criteria for compulsive sexual behavior (CSB) should be included in the new International Classification of Diseases (ICD-11) [24], and this proposal has been accepted. According to the upcoming ICD-11's diagnostic criteria for impulse control disorder [25], patterns of failure to control intense sexual impulses or urges and the resultant repetitive sexual behaviors are considered to be the characteristic features of the disorder. The BPS considers compulsive pornography to be the core component of problematic pornography use. Moreover, the BPS has been used with different samples, and it has demonstrated satisfactory psychometric properties among American and Polish pornography users [26]. Many past studies have used the BPS to identify pornography addicts. Furthermore, it has also been used to ascertain the severity of problematic pornography use among men who seek pharmacologic or psychological treatment as a result of their loss of control over their sexual behaviors [27][28][29]. Therefore, in this study, the BPS scores were used as the reference standard against which the sensitivity and specificity of the three aforementioned scales were ascertained.
Several recent reviews have focused specifically on the conceptualization and assessment of problematic pornography use [4,11,30,31]. Some reviews have briefly summarized and commented on the included instruments [5], whereas others have evaluated their ability to assess the core components of problematic pornography use [11]. However, no past study has compared the different scales and identified the most accurate measure of problematic pornography use using a same standard or indicator. Measures of problematic IPU are heterogeneous, and each scale focuses on a different aspect of problematic IPU. Furthermore, because these scales have not been extensively validated, it is difficult to compare the findings of the studies that have used them. In addition, the sensitivity of the different scales that assess problematic IPU have not been adequately compared. Therefore, in the present study, a QUAN→QUAL mixed-methods design was conducted, including (1) using quantitative methods to identify a scale with a higher sensitivity index from three selected scales (PPCS, PPUS, s-IAT-sex) for assessing problematic IPU. Moreover, the duration of usage, frequency of engagement in OSAs, sexual compulsivity, and pornography cravings were used to examine the criterion validity of the assessments. Subsequently, (2) qualitative interviews were conducted with volunteers and therapists who have serviced the individuals in trouble of problematic IPU to further examine the appropriateness of the "more accurate" scale from the service providers' perspectives, whereby the qualitative part helps to evaluate and interpret the results obtained from the main quantitative study.

Sample
The study sample consisted of 560 men and 412 women, and the mean age of the sample was 24.8 years [standard deviation (SD) = 7.2 years; range = 18-48 years]. The group comparisons of the demographic characteristics of the three study samples can be inferred from Table 1.

Instruments
Three Main IPU Measurements PPUS. The PPUS is a 12-item self-report scale that assesses four dimensions of IPU [15]: distress and functional problems, excessive use, difficulties in self-control, and IPU to escape or avoid negative emotions. In the Chinese version of the assessment, the term "pornography," which was used in the original scale, was modified as "internet pornography" in all instances (e.g., "I spend too much time being involved in thoughts about internet pornography"). The participants were required to indicate the frequency with which they had engaged in IPU during the past 6 months on a six-point scale that ranged from 0 (never) to 5 (all the time). Higher scores were indicative of a greater severity of engagement in IPU. The Cronbach's alpha of the total scale was 0.95 in this study.
PPCS. The PPCS was used to measure problematic IPU [14]. Responses were recorded on the following 7-point scale: 1 = never, 2 = rarely, 3 = occasionally, 4 = sometimes, 5 = often, 6 = very often, 7 = all the time. PPCS consists of 18 items, and assesses the six core components of addiction: salience, mood modification, conflict, tolerance, relapse, and withdrawal. Each factor is measured by three items (e.g., "I felt that I had to watch more and more internet porn for satisfaction" is an item of measure "tolerance"); the Cronbach's alphas of the aforementioned six factors were 0.77, 0.84, 0.71, 0.78, 0.86, and 0.86, respectively, in the study. The Cronbach's alpha of the total PPCS was 0.96. A cutoff score of 76 was used to ascertain normal and problematic use; specifically, scores that were greater than 76 were indicative of problematic use.
s-IAT-sex. Responses to each of the 12 items of the s-IAT-sex are recorded on a five-point scale that ranges from 1 (never) to 5 (always) [9]. The scale consists of two dimensions. The first factor assesses poor self-control and difficulties in reducing the amount of time that is spent online (six items, e.g., "How often do you find that you stay on Internet sex sites longer than you intended?"), whereas the second factor measures the functional impairments that are associated with engagement in cybersex (six items, e.g., "How often do you feel depressed, moody, or nervous when you are offline, which goes away once you are back on internet sex sites?"). The composite score, which can be computed by summing the individual item scores, can range from 12 to 60; higher scores are indicative of greater problems. The internal consistency (Cronbach's alpha) coefficients of the total scale and first and second factors were 0.89, 0.77, and 0.88, respectively, in this study.

Criterion Validity Questionnaires
PCQ. This 12-item questionnaire is a unidimensional assessment [32,33]. The following are a few sample items: "If the situation permitted, I would watch pornography right now" and "If I were to watch pornography right now, I would have difficulty stopping." The respondents were required to indicate how strongly they agreed with each item using the following seven response options (presented without numerals): "completely disagree," "somewhat disagree," "disagree a little," "neither agree nor disagree," "agree a little," "somewhat agree," and "completely agree." Higher scores are indicative of a greater craving for pornography. The Cronbach's alpha of this scale was 0.92 in the current study. The instructions of the PCQ present a craving-for-pornography vignette, which requires the respondent to imagine that they are alone in their room and seated in front of their computer and that they have a strong urge to watch their favorite type of pornography.
The Sexual Compulsivity Scale (SCS). The extent to which participants exhibit the characteristics of compulsive pornography use was assessed using the 10-item SCS that has been developed by Kalichman et al. [34]. Responses were recorded on a four-point rating scale (1 = not at all like me, 2 = slightly like me, 3 = mainly like me, 4 = very much like me, e.g., "I have to struggle to control my sexual thoughts and behavior"). In this study, the Cronbach's alpha of this scale was 0.86.
Questionnaire of OSAs. Thirteen items were used to measure participants' use of the internet for the following purposes: (1) viewing sexual explicit materials (SEM), (2) seeking sexual partners, (3) cybersex, and (4) flirting and sexual relationship maintenance [35]. Viewing SEM was assessed using five items (e.g., visiting erotic/pornographic websites, viewing and downloading erotic/pornographic videos from the internet, reading erotic/pornographic material online), each of which required responses to be rated on a nine-point scale that ranged from 1 (never) to 9 (at least once a day). The other three subscales assessed frequency using a nine-point scale that ranged from 1 (0 times) to 9 (20 or more times). Two items measured the frequency with which the respondents had sought sexual partners as well as the number of sexual partners that they had sought and found online. The frequency of engagement in cybersex was assessed using four items (e.g., masturbating or viewing strangers masturbating in front of a webcam, describing sexual fantasies either through texts or orally). Internet use for the purposes of flirting and sexual relationship maintenance was measured using two items. The Cronbach's alpha of the entire scale was 0.88 in the study. Higher scores were indicative of more frequent engagement in OSAs.
Additional Questions about IPU. In addition to items that assessed demographic characteristics, a few questions that were related to IPU were also posed to the participants. After providing them with a clear definition of internet pornography, the participants were asked to indicate their age of first exposure to pornography and the duration of time that they typically spent watching internet pornography every week.

The Reference Standard-BPS
The BPS, which has been developed by Kraus et al. [26], was used to assess pornography use during the past 6 months. This five-item assessment uses a three-point rating scale (0 = never, 1 = occasionally, 2 = always, e.g., "You find it difficult to resist strong urges to use sexually explicit material."); a cutoff score of 4 was used to detect problematic pornography use (absolute range = 0-10). Higher scores are indicative of more problematic pornography use. The Cronbach's alpha of the BPS was 0.84.

Procedure
This online study was conducted through a popular Chinese survey website, namely, Wenjuanxing (www.sojump.com). Adult members of the website received an email with a link that redirected them to the survey website and a brief introduction to our survey. This brief introduction informed the recipients that they were eligible for participation if they had engaged in IPU during the past 6 months (e.g., reading online pornographic content, browsing pornographic websites, sharing/watching pornographic videos or pictures, interacting and flirting with others) and were interested in participating in the survey. A total of 972 valid responses were collected from participants from 110 cities in 28 of the 34 provinces/regions in China (i.e., identified using the internet protocol addresses). As expected, all participants obtained scores that were equal to or greater than 14 on the measure of OSAs (the lowest possible score is 13, and it indicates no prior IPU); this indicated that all of them had engaged in at least one OSA during the past 6 months. Three highly homogeneous samples were required to respond to the three measures of problematic IPU, namely, the PPCS, PPUS, and s-IAT-sex, respectively. Each sample also completed the aforementioned mentioned assessments against which their criterion validity was to be examined. This study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of the Department of Psychology, Fuzhou University (date of approval, 7 April 2019).

Analysis
Statistical analyses were conducted using SPSS 19.0 (IBM, Armonk, NY, USA) and Mplus version 7 [36]. Item-total correlations were computed to identify items that functioned poorly. CFA was used to test the factor structures of the scales of interest. Maximum likelihood estimation with the Satorra-Bentler correction was used to determine the fit between the data and the factor structures. Model fit was tested by inspecting the following indices: root mean square error of approximation (RMSEA; good: ≤0.06, acceptable: ≤0.08), comparative fit index (CFI; good: ≥0.95, acceptable: ≥0.90), and Tucker-Lewis index (TLI; good: ≥0.95, acceptable: ≥0.90). The reliability of the scales was assessed by computing Cronbach's alpha coefficients.
To identify possible groups of at-risk pornography users, latent profile analysis (LPA) was used. LPA was conducted using the original dimensions of each scale as explicit variables, and different groups of individuals with problematic IPU were successively divided into two to four categories for model fitting estimation. Sensitivity was defined as the proportion of persons with positive symptoms (as detected by the BPS) and members of the at-risk group (identified through LPA), whereas specificity was defined as the proportion of persons with negative symptoms and the nonproblematic group [37].

Validation of the Three Scales
The results of item analysis, CFA, and tests of reliability and convergent validity are shown in Table 2. Item-sum correlations were computed to examine item functioning. The PPCS and PPUS yielded higher coefficients, and both these scales also yielded good fit indices (i.e., CFA) and stronger reliability coefficients. PPCS, PPUS, and s-IAT-sex significantly positively related with SCS, PCQ, OSAs and usage time severally, and PPCS demonstrated stronger convergent validity.

LPA
The results of LPA are shown in Table 3. For PPCS, the Lo-Mendell-Rubin adjusted likelihood ratio test (LMRT) results were significant when the number of classes was 4, and the entropy value was lower. Thus, the classification accuracy was not as high as that of the three-class solution; accordingly, the three-class solution was selected. For PPUS, when the model consisted of three classes, the LMRT results were significant; furthermore, the entropy value was evidently higher than that of the four-class solution. With regard to the s-IAT-sex, the nonsignificant p-value that emerged for the LMRT results suggested that the three-and four-class solutions should be rejected in favor of the two-class solution. With regard to the three groups that emerged for the PPCS and PPUS, the first class obtained the lowest averages across all the scale dimensions; thus, this group was referred to as nonproblematic consumption. The second class obtained moderate scores on all the scale dimensions; therefore, these group members were referred to as low-risk pornography users. The third class obtained the highest scores on all the scale dimensions; thus, this group was referred to as at-risk users. As shown in Table 4, with regard to the two classes that emerged for the s-IAT-sex, class 1 obtained lower scores than class 2 on both the scale dimensions; therefore, they were referred to as the nonproblematic and at-risk groups, respectively (group differences in scores on the specific dimensions are shown in Appendix A).  3 s-IAT-sex = Short Internet Addiction Test Adapted to Online Sexual Activities, 4 Sensitivity = the proportion of persons with positive symptoms and members of the at-risk group that was identified through LPA, 5 Specificity = the proportion of persons with negative symptoms and the nonproblematic group.

Sensitivity and Specificity Analysis
The results showed that the sensitivity of the PPCS was 89.66%, which is higher than the values that emerged for the PPUS (i.e., 81.25%) and the s-IAT-sex (i.e., 71.72%). There were differences in the specificity of the three scales, and the values ranged from 85.86% to 94.95%. The PPCS demonstrated greater sensitivity (89.66%), and its specificity was 85.86%. This indicates that approximately 10% of problematic users had been classified as nonproblematic users and that approximately 14% of nonproblematic users had not been identified. In general, the PPCS and PPUS performed better than the s-IAT-sex. Since this study aimed to identify the scale with greater sensitivity in detecting problematic IPU, the PPCS was investigated in greater detail.

Sample
We interviewed 22 (20 men; mean age = 27.2) problematic IPU service volunteers (who provide online services on the following website: http://www.ryeboy.org/; average service time = 3.3 years) and 11 therapists (who have worked with individuals with problematic IPU and had more than 3 years of clinical experience).

The Interview Outline
Since the used scales were easy to administer and consisted of close-ended questions, interviews were conducted to examine participants' perspectives more deeply and comprehensively. The interview guide primarily sought to explore interviewees' understanding of problematic IPU/addiction and their evaluations of the dimensions of the selected scale. The interviewees were required to rate the importance of the dimensions on a scale that ranged from 1 (not at all important) to 7 (very important).

Procedure
In this study, we primarily explored their understanding of the concept of problematic IPU and the dimensions of the recommended scale. Two psychology graduate students served as the interviewers. At the beginning of the interview, the interviewees were informed about the purpose and significance of the interview and assured of the anonymity and strict confidentiality of their interview data; the interviews were recorded with their permission.

Analysis
The interview recordings were transcribed into verbatim scripts, and participants' identifying information was concealed. Next, we undertook thematic analysis of the text; in other words, we collated different interviewees' responses to the same question to create new text. Tree Nodes were established based on the dimensions of the selected scale, and interviewees' original statements were identified and summarized as a named code. Through this process, NVivo automatically generated statistics for all the references of the texts.

Results
With regard to the characteristics of problematic IPU, we generated a total of 20 codes by analyzing the interview data. Among these features, preoccupation with IPU (22 mentions), IPU to escape or avoid a negative emotional state (21 mentions), interpersonal conflict (22 mentions), and physiological and psychological symptoms (45 mentions) were most commonly mentioned. Furthermore, the 20 codes were summarized into the six dimensions of the PPCS (see Figure 1).

General Discussion
Problematic IPU is still a controversial issue; notably, it appears that no real consensus exists regarding the conceptualization and screening tool of problematic IPU. Several scales are available; thus, the assessment of problematic IPU is inconsistent, indicating that findings in this area are not readily comparable. The present study aimed to selected a more sensitive scale to screen problematic IPU, because higher sensitivity implies lower rate of missed diagnosis (i.e., problematic users who have been incorrectly screened as nonproblematic users). Basing on a systematic literature review, three scales were retained. Considering that research with mixed methods combining quantitative and qualitative analyses can enrich and improve our understanding of complicated phenomena [38,39], a quantitative method was used to identify a "more accurate" analysis from the three retained scales. Results of CFA showed that all three scales have good applicability in the wide range of adult groups (age in this case ranged from 18 to 45 years) in three highly homogeneous samples; compared to the other two scales, the PPCS demonstrated greater sensitivity and comparative specificity among samples drawn from the general population (results of the QUAN). Considering that the expression of questionnaire survey is brief and closed, and that the interview can understand the participants' undefined views more deeply and comprehensively, subsequently, results of QUAL showed that symptoms of problematic IPU proposed by the servers (volunteers and therapists) can be grouped into the six dimensions of PPCS and most of the servers supported the six-factor structure of PPCS.

Instance of the interview:
Interviewer: According to your service experience, what do you think is problematic internet pornography use? In other words, what are the expressions/symptoms of problematic internet pornography use? Interviewee (service volunteer): They (problematic users) show difficulty controlling the craving for internet pornography (code: pornography carving), they are unable to control their own behavior, for instance, browsing pornographic websites, masturbating while watching porn frequently (code: difficulties in control). Their brains are constantly bombarded with sexual materials (code: preoccupation). If they are not exposed to internet pornography, they will feel uncomfortable, or feel that their heart is empty (code: depression resulting from unsuccessful withdrawal).
After presenting interviewees with the definitions of the six components of problematic IPU and further clarifying their meaning using examples, we presented them with questions "Based on your service experience, do you endorse this structure? Which dimension or dimensions do you think are particularly central to IPU?" Most (>95%) participants endorsed the six dimensions. It also can be inferred from Figure 1 that both volunteers and therapists emphasized the centrality of conflict, relapse and withdrawal in IPU (basing the frequency of mentions); at the same time, they weighted the mood modification, relapse and withdrawal as more important features in the problematic use (basing the important rating).

General Discussion
Problematic IPU is still a controversial issue; notably, it appears that no real consensus exists regarding the conceptualization and screening tool of problematic IPU. Several scales are available; thus, the assessment of problematic IPU is inconsistent, indicating that findings in this area are not readily comparable. The present study aimed to selected a more sensitive scale to screen problematic IPU, because higher sensitivity implies lower rate of missed diagnosis (i.e., problematic users who have been incorrectly screened as nonproblematic users). Basing on a systematic literature review, three scales were retained. Considering that research with mixed methods combining quantitative and qualitative analyses can enrich and improve our understanding of complicated phenomena [38,39], a quantitative method was used to identify a "more accurate" analysis from the three retained scales. Results of CFA showed that all three scales have good applicability in the wide range of adult groups (age in this case ranged from 18 to 45 years) in three highly homogeneous samples; compared to the other two scales, the PPCS demonstrated greater sensitivity and comparative specificity among samples drawn from the general population (results of the QUAN). Considering that the expression of questionnaire survey is brief and closed, and that the interview can understand the participants' undefined views more deeply and comprehensively, subsequently, results of QUAL showed that symptoms of problematic IPU proposed by the servers (volunteers and therapists) can be grouped into the six dimensions of PPCS and most of the servers supported the six-factor structure of PPCS.
Among the three scales, the PPCS score was most robustly related to the duration of usage, frequency of engagement in OSAs, and pornography cravings. Problematic IPU can appear under the umbrella of hypersexuality similarly to frequent engaging in various forms of cybersex, intense craving for pornography, and compulsive sexual behaviors [40], insofar that the robust relationship not only demonstrated a higher criterion validity, but also implied that co-screening instruments (i.e., pornography craving, frequency and duration of use, compulsive use) are expected to work as auxiliary screening indicators. Recent studies have revealed that for some people, pornographic use gave rise to their feeling of discord and shame contributing to their conflict of actual sexual materials consumption and their belief; in turn, these feelings of distress and shame may drive a morbid self-perception that they are addicted, but this may not be a real behavioral disorder [41,42]. In order to avoid misjudgment due to the self-perceived problematic use, it is more advisable to combine other supporting scales, and the combination diagnosis indexes of the diversity were selected to screen the prevalence of problematic IPU. In this study, with the higher correlation of PPCS with frequency of OSAs, the PCQ showed that combined with other indicators, it can better screen out problematic use and is more likely to avoid the misjudgment caused by subjective self-perceived addiction.
The more robust psychometric properties and higher recognition accuracy of the PPCS may be attributable to the fact that it has been developed in accordance with Griffiths's six-component structural theory of addiction (i.e., in contrast to the PPUS and s-IAT-sex). The PPCS has a very strong theoretical framework, and it assesses more components of addiction [11]. In particular, tolerance and withdrawal are the important dimensions of problematic IPU that are not assessed by the PPUS and s-IAT-sex; PPCS is the only instrument that explicitly assesses the "tolerance" component [11,14]. According to the "two-phased" internet pornography addiction model, in which the first step is characterized by an excessive use to internet pornography, and the second functions as a marker by repeated failures to break free from excessive use, despite negative consequences [43]. Items related to information about salience, carving, and tolerance reflect the engagement in internet pornography, corresponding to the first step, whereas items related to withdrawal, relapse, and conflict measure addiction more, corresponding to the second step. Obviously, components of PPCS includes both engagement in pornography and addiction of IPU, which has an intact theoretical framework of addiction.
The PPCS appears to be a more valid instrument for assessing problematic pornography use, has potential application in detecting prevalence concerning problematic IPU or cybersex addiction, and may be useful in assessing treatment outcomes. Our findings indicate that individuals who score high on the PPCS also report frequent engaging in various forms of online sexual activities, intense craving for pornography, and compulsive sexual behaviors. Thus, it appears important for clinicians to be aware of problematic pornography use and its related associations such as pornography craving, compulsive use. Moreover, it is important to note that the scale PPCS is recommended as a screening instruments to identify problematic users in the public and assess the prevalence rather than a diagnostic tool; future studies should further research its validity and cutoff in clinical sample; we also encourage individuals to visit a clinical therapist after being identified with problematic IPU by the use of PPCS.
This study has several limitations. First, data were collected using self-report measures; therefore, the reliability of the results depends on the respondents' honesty and accuracy of their comprehension of the scale items. Second, the study sample was recruited through an online survey company; therefore, the participants of this study may have been more educated and affluent than the average Chinese person. Furthermore, the study participants primarily lived in the capital/provincial capital, cities, and towns. Third, because the sample consisted of only a small number of non-heterosexual subjects, it was not possible to examine whether the factor structure and meaning of the contents of the PPCS differed across individuals with different sexual orientations.

Conclusions
The present study showed that the PPUS, PPCS, and s-IAT-sex are promising measures of problematic IPU. However, when sensitivity and specificity were simultaneously examined, the PPCS emerged as a more suitable measure of problematic IPU. The qualitative findings further confirmed that service providers endorsed the underlying structure of the PPCS. Funding: This work was supported by the National Social Science Foundation of China (Grant No. CEA150173 and 19BSH117) and the education reform project of Fujian province (FBJG20170038). The funding agencies did not have input into the content of the manuscript and the views described in the manuscript reflect those of the authors and not necessarily those of the funding agencies. non-governmental organization focusing on helping the problematic internet pornography users) for their help to recruit volunteers who serviced the addicts in the qualitative step, and pay tribute to them for their effort in helping problematic users.

Conflicts of Interest:
The authors report no conflict of interest with respect to the content of this manuscript.
Appendix A Figure A1. The average scores of the three latent classes based on the dimensions of PPCS. Note: PPCS = Problematic Pornography Consumption Scale, range = 1-7; *** p < 0.001 indicate that the score of at-risk group was significantly higher than that of low-risk group; △△△ p < 0.001 indicate that the score of low-risk group was significantly higher than that of non-problematic group; ◇◇◇ p < 0.001 indicate that the score of at-risk group was significantly higher than that of non-problematic group. The same below.  of at-risk group was significantly higher than that of low-risk group; p < 0.001 indicate that the score of low-risk group was significantly higher than that of non-problematic group; p < 0.001 indicate that the score of at-risk group was significantly higher than that of non-problematic group. The same below.

Conflicts of Interest:
The authors report no conflict of interest with respect to the content of this manuscript. Appendix A Figure A1. The average scores of the three latent classes based on the dimensions of PPCS. Note: PPCS = Problematic Pornography Consumption Scale, range = 1-7; *** p < 0.001 indicate that the score of at-risk group was significantly higher than that of low-risk group; △△△ p < 0.001 indicate that the score of low-risk group was significantly higher than that of non-problematic group; ◇◇◇ p < 0.001 indicate that the score of at-risk group was significantly higher than that of non-problematic group. The same below.