Sunday, February 7, 2016

Internet Addiction: A Critical Comparison of Two Articles - Paper for NCU PSY7102






Internet Addiction:  A Critical Comparison of Two Articles
Laura Ann Collins
Northcentral University


Abstract
With the development of multiple ways to connect to the internet and the importance of doing so in modern life, concerns have arisen about pathological internet use – often described as internet addiction (IA).  The body of literature exploring concerns about IA is rapidly expanding.  Conceptualization of the disorder is evolving as researchers seek to determine the definition of IA and develop tools to identify IA with accuracy and precision.  This article is a critical comparison of two recent research articles on IA.  The first article reviews existing scales which measure internet addiction to illuminate strengths and weaknesses of the scales (Laconi, Rodgers, & Chabrol, 2014).  The second article is a meta-analysis in which the authors seek to measure prevalence of IA and link prevalence to experienced quality of life (Cheng & Li, 2014).  Each article will be briefly reviewed.  Common themes will be discussed.  Finally, future areas of research will be highlighted.
Keywords:  Internet Addiction, IA


Internet Addiction:  A Critical Comparison of Two Articles
Over the last two decades (Laconi et al., 2014), there has been increasing concern about internet use.  Questions regarding what is appropriate internet use, when internet use becomes pathological, and who is most at risk for pathological internet use have become topics for research.  Through use of Roadrunner search engine which searches multiple research databases, the first identified scholarly use of “internet addiction” was in 1983.  The next identified scholarly use was in 1996 article when OReilly interviewed Dr. Kimberly Young and then suggested a “new disorder” entered “the medical lexicon” (OReilly, June 15, 1996).  OReilly also mentioned that Dr. Young had “put together” an assessment tool and used established criteria for addictions to create/align criteria for internet addiction.   However, the criteria OReilly describes as “standardized” by Young is anything but standardized.  Instead, the structure and details which comprise internet addiction is being debated and developed through ongoing research. 
Of essential importance to this debate is understanding the psychometric properties of the tools used to conduct the research.  A tool which establishes correlation is much different than a tool which establishes causation.  Both types of information can be valuable – but only when the researcher understands which piece of information the tool provides.  It is for this reason that the first article selected for critical comparison is a review of commonly used IA assessment tools.. 
Once tools are validated, it is then important to understand how the tools can be useful to researchers.  The purpose of research is to understand a phenomenon, and if able, to predict future occurrence.  The second article was selected to provide an example of use of assessment, incorporation of data derived from the assessment into theory, and how theory can influence understanding of phenomenon.
Further, it is important that each researcher understand the limits of the tool/assessment which he or she uses.  The researcher which does not understand limitations of tools is at risk of too broadly applying results of research, thereby introducing error into the body of literature.  It is through integrating the information from the first article into the second that the research can be critically reviewed and avenues for new research may be seen.

Article One:  The measurement of internet addiction: A critical review of existing scales and their psychometric properties.  By Laconi, Rodgers, and Chabrol

Laconi et al. (2014) determined that there was a surfeit of IA assessment tools, but limited information about the validity and usefulness of those tools.  Therefore, they conducted several searches of scholarly article databases in order to identify scales measuring IA.  They identified 92 papers pertaining to IA scales. From these papers, 45 assessment tools were identified.  Seventeen scales had more than 1 study validating the scale.  Ten scales had 3 or more studies providing validation.  Each scale, if possible, was evaluated based on published research for 1) reliability, 2) validity, and 3) factor structure.  Each scale’s strengths and weaknesses were discussed.
Of particular interest was that the factor structure created by factor analyses of each scale had significant variation.  Per Laconi et al., the lack of factor structure may be rooted in the lack of definition and lack of theoretical grounding leading to “lack of construct validity” (2014, p. 196).  Further, they call for a cessation in development of new scales, with focus instead being on robust validation of existing scales.
The scale which Laconi et al. (2014) consider the most frequently used and the most validated was the Young Internet Addiction Test (IAT).  However, they expressed concerns about the lack of validated cut-off scores, outdated or vague items, and the variation in reliability based on which ethic group is being tested.  They quote Faraci et al. (2013) as describing the IAT as lacking “rigorous and systematic psychometric evaluation” (Laconi et al., 2014, p. 198).  Laconi et al. call for cross-cultural research, utilizing large sample sizes and paying special attention to the psychometric properties of the assessment tools in order to properly validate the tools.

Article Two:  Internet addiction prevalence and quality of (real) life: A meta-analysis of 31 nations across seven world regions.  By Cheng and Li

Cheng and Li (2014) sought to provide world-wide prevalence data for IA through analyzing research articles from 1996 to 2012 in which the Young Internet Addiction Test (IAT) or the Young Diagnostic Questionnaire (YDQ) was utilized to assess for IA.  They then attempted to determine if this data correlated with experienced quality of life as measured through national indices and The Life Satisfaction Index in order to determine if either the accessibility hypothesis or the quality of (real) life (QOL) hypothesis fit the data.
Cheng and Li state that as of “December 2013, approximately 39% of the world’s 7 billion people use the internet” (2014, p. 755).  Rate of use per population varied between North America at 85% and 16% in Africa.  They searched databases between 1996 and 2012, resulting in 80 acceptable articles which yielded 164 independent samples, averaging 554 participants per sample, and including 31 different nations (2014, p. 758). 
They determined an IA global prevalence rate of 6%, which they noted was 3 times higher than the global prevalence of pathological gambling.  They determined that IA prevalence negatively correlated with decreased life satisfaction and national income, and correlated positively with increased pollution and commute time.  These results, they posited, are tentative support for the QOL hypothesis of IA. Limitations described by Cheng and Li (2014) include lack of data for African nations, need for longitudinal studies to determine incidence and remission rate, and additional determination as to which hypothesis accurately predicts IA prevalence. 

Discussion:  Common Themes, Integration, and Critical Comparison

The IAT was developed by Young in order to track what she saw as an emerging problem (OReilly, June 15, 1996).  Since then, it has become one of the most utilized tools to assess IA (Laconi et al., 2014).  The popularity of this test can be seen in Cheng and Li’s selection of research articles utilizing this test.  Given that both studies analyze research conducted during approximately the same time period, limitations of the IAT are significant.
Unfortunately, Cheng and Li do not mention this concern in the section of their report which details limitations. Just one concern identified by Laconi et al. (2014) is the reliability of the IAT for groups other than Asian participants or high school students is a concern.  Multiple other concerns were identified by Laconi et al. (2014).  Secondly, Cheng and Li do not discuss limitations of the assessment tools which are self-report instruments.  Finally, Cheng and Li do not detail the primary assumption upon which their research is based… namely that Young’s conceptualization of IA is accurate.  However, OReilly (June 15, 1996) detailed Young’s process of development of IA criteria as taking the criteria for substance addiction and substituting the word internet for substance. 
Cheng and Li’s method of determining IA structure presupposes that IA is identical in structure to substance addiction.  However, Laconi et al. (2014) provide data (factor analyses) which suggests the IAT does not comprehensively assess structure of IA. Cheng and Li (2014) selected research which used a specific assessment instrument, adopted the criteria for IA upon which development of those assessments were based, and then made assumptions on analyses of the selected data.  These assumptions were applied to two hypotheses, with the result being one hypothesis was supported while the other was not.  No discussion of the multiple limitations of the IAT illuminated by Laconi et al. (2014) was provided.

Conclusion

Both studies are valuable contributions to the body of literature.  Laconi et al. (2014) analyze multiple IA assessment tools, suggest areas where the tools can be improved, and detail strengths of the tools.  Cheng and Li (2014) analyze multiple research papers to determine global prevalence of IA and to attempt to determine why IA might occur.  Unfortunately, the articles are not equally valuable.  Laconi et al. (2014) are quite careful to describe possible limitations of their work, limitations of the tools, and to call for additional research.  Cheng and Li (2014) do not seem to be as careful in describing the limitations of their work and do not describe any limitations of the tools upon which the research they analyzed is based.  They do call for further research.  However, there is a difference between Laconi et al.’s call for research to validate instruments (non-specific result) and Cheng and Li’s call for additional research to further validate the hypothesis they find their data supported (specific result requested).
When reading a research article, understanding the limitations of the tools is essential.  When conducting research, reporting the results must include discussion of the limits of the tools and the research should be reported.  Likewise, assumptions upon which the research is based should be delineated.  Without reporting the limits and assumptions which permeate the research, the significance of the research is unable to be determined by the reader.  This lessens the value of the research to the body of literature and to the advancement of science.
Future research regarding IA needs to develop robustly validated tools for assessment and consensual agreement regarding criteria and definition of IA.  Until these two foundational items are met, research determining international prevalence is limited in usefulness.  Additionally, future research, regardless of topic, should be careful to explicitly tabulate limitations in order to provide the reader with information to establish relevance and value to the extant body of literature.



References

Cheng, C., & Li, A. (2014). Internet addiction prevalence and quality of (real) life: A meta-analysis of 31 nations across seven world regions. Cyberpsychology, Behavior, and Social Networking (17) 12, 755-760. doi: 10.1089/cyber.2014.0317.
Laconi, S., Rodgers, R., & Chabrol, H. (2014). The measurement of internet addiction: A critical review of existing scales and their psychometric properties. Computers in Human Behavior, 41, 190-202. doi: 10.1016/j.chb.2015.04.056.
OReilly, M. ( June 15, 1996). Internet addiction: A new disorder enters the medical lexicon. CMAJ: Canadian Medical Association Journal = Journal De L'association Medicale Canadienne [serial online].154(12), 1882-1883.




No comments:

Post a Comment