The Rhetorist

Increasing growth and diffusion of the Internet is testimony to the fact that more people are getting on-line and, once on-line, staying there. Some researchers such as Dr. Kimberley Young of the University of Pittsburgh at Bradford have suggested that this is indicative of a psychological problem not unlike the predilection certain individuals may have toward substance abuse.

Any screening for possible Internet addiction as substance abuse should include evaluation of the user’s Internet connectivity habits. This evaluation may reveal a number of habitual patterns similar to persons who are addicted to substances such as alcohol, narcotics or gambling. One of the expected findings may be an individual’s inability to disconnect from the Internet, even though harm may be evidently occurring as a result of steady connectivity. Such harm may be evidenced as a physical ailment, a missed appointment, a missed meal or even a missed mortgage payment because instead of paying the rent the individual is paying an on-line service to stay frequently connected. Other findings may indicate that users feel let down when disconnected, or feel a constant urge to reconnect. Still other findings may reveal patterns or instances of previous addiction in a user’s genealogy.

Some scholars have labeled this condition Internet Addiction Disorder (Goldberg, 1995). However, research has systematically attempted to build a typology of Internet-related disorders. Therefore words like “addiction” and “dependency” are not as distinctive in this realm as they are in the realm of more tangible substance abuses such as alcohol, gambling or narcotics. This study makes a preliminary attempt to determine usage patterns among self-classified addicts and dependents, and attempts for the first time, a classification scheme (reclassification) based on usage frequencies and perceived consequences.

Dr. Morton Orman claims that, when confronting Internet addiction, it is “no different than dealing with any other type of addiction. Whether you are addicted to heroin, gambling, cigarettes, sexual deviancy, or eating Milky Way bars, all addictions have certain basic elements in common.” Orman, author of “The 14 Day Stress Cure,” has mapped out the phases an individual confronting Internet addiction probably goes through. Beginning with denial, the progression leads to an individual’s failure to ask for help. Lack of other pleasures; underlying deficiencies in coping and in life management skills; giving in to temptation; failing to keep one’s word; and, failing to do what may be necessary, are all vital components of Orman’s strategy for coping with this addiction. He wraps up his analysis with the individual’s failure to anticipate and deal with the possibility of relapses.

Given the exponential growth of the Internet and today’s ever-increasing ease of access, Internet connectivity may be giving rise to exponential growth of Internet addiction. But ease of access or connection does not necessarily mean ease of disconnection. Neither is an excessive amount of time on-line necessarily indicative of addiction or dependency. People are daily being given new reasons and incentives to connect to the Internet, and, once connected, to stay there. In spite of whatever reason or combination of reasons they may be using to connect to the Internet, many of these people seem to be realizing that their time on-line may be disproportionate with their allocation of time for other daily habits and routines.

Accessing the Internet for e-mail, stock quotes, employment opportunities, cybersex and the further acquisition of knowledge are all among the most common reasons given for increased Internet connectivity. New technological advances help make access to the Internet a habitual process for many people today. Consequently, many individuals are suddenly realizing that they have made their connection to the Internet a priority in their lives, and that there may be more than just a monetary price attached to their habit.

This has led to the assumption on the part of some Internet users that they may be addicted to or dependent upon this communication medium. This study focuses on differences between those persons who appear to be addicted to the Internet and those persons who appear to be dependent. This researcher takes the position that there are differences between these two types of Internet users, and that there are consequences which differentiate the two groups in accordance with their usage patterns, as well as similarities which share equal validity. Analysis consists of evaluations of survey data as it relates to these commonalities and differential data patterns as they are relevant to the above hypothesis.

Method

The method used for this research project was partially indicated by immersion of the researcher into the studied environment as a fellow addict or dependent. Of major importance to the research was the development of a new tool which would adequately provide for measurement of any important variables and yet interface with the actual medium being studied. Consequently, I developed a survey form (McSurvey) and placed it on-line for access by participants. Subjects were invited to participate via an e-mail invitation addressed to them from sites where they had publicly indicated addiction, dependency or use of the term “hooked” for their personal Internet connectivity habits.

Addiction Sites

Without doubt, the majority of queried individuals who eventually became participants came via sites professing addiction. Some of these sites are designed to be humorous places to post personalized comments about unusual computer habits, while others profess to provide relief for genuine addiction. Almost all sites yielded a ripe source of potential subjects who were given the opportunity to further discuss and comment on their addictions via specific comment fields on each site. Some of the sites visited were Webaholics Anonymous, World Wide Webaholics, Interneters Anonymous, Netaholics Anonymous, Transformations and the Webaholics Support Group.

E-mailing people from sites where addiction was noted in their comment fields, or where they may have been listed as charter members of an addiction support group, became the primary method of invitation. There were also other means which became exciting alternatives to this procedure and yielded participants. Gregory Collins, as president of Webaholics Anonymous, was very interested in this research project and graciously attached a banner at a highly prominent point on his site which linked visitors to the McSurvey. Transformations allowed for posting on their bulletin boards. Again, this turned out to be a very worthwhile way for visitors of these sites to link to the survey, as some did.

Definitions

Respondents were asked to respond to over 30 questions, most of which queried their Internet connectivity habits. Questions 26a and 26b were both designed to reveal the definitions which lie at the core of this study. Addiction and dependency are distinguished in Question 26a as the following:

• Addiction implies that you may have no will of your own concerning a particular habit.
• Dependency implies that you have a strong, compelling desire to continue a particular habit.

Other scholars have suggested that the term “to be hooked [on computers]” implies “the essence and condition of dependency,” but “avoids the very negative, emotive allusions” (Shotton, 1989) made by other words such as addiction, and, of course, dependency. This observation was taken into consideration when initially searching for participants with the specific keyword choices, and also when constructing these definitions for addiction and dependency.

Many participants believed that their connectivity habits had led them to a negative state of affairs in their lives, and this is evidently the expectation of those who consider themselves either addicted or dependent. Nonetheless, in spite of any battle over the specific peculiarities of these word choices, one must admit that usage of any of them would probably have the capacity to make a survey respondent a bit uncomfortable no matter what the study environment or scenario.

CAGE Model

The primary factor used to develop the independent variable was the CAGE model. CAGE is an acronym used by many psychologists to provide evidence of possible addiction to a substance. The CAGE model is used extensively for screening possible alcohol and/or other drug abuse. McSurvey questions 4-7 implemented the CAGE model, as well as 15-21, the latter of which further affirmed Questions 4-7 and provided more potential for in-depth analysis of positive responses which were expressed through them. The model was initiated in the following way:

• 4a. Have you ever felt that you should cut down on your Internet connectivity?
• 5b. Have people annoyed you by criticizing your Internet connectivity habits?
• 6a. Have you ever felt bad or guilty about your Internet connectivity?
• 7a. Have you ever connected to the Internet early in the morning?

I first learned of the CAGE model from Professor Lynn Kozlowski, head of the Department of Biobehavioral Health at Pennsylvania State University, who suggested that I adapt a working CAGE model from a source screening for substance abuse. I adapted a CAGE model as a screening method for McSurvey respondents based on The Physician’s Guide to Helping Patients with Alcohol Problems, an on-line guide prepared by the National Institute on Alcohol Abuse and Alcoholism which is published by the National Institutes of Health.

Internet Addiction Disorder

No current research on the subject of Internet addiction could be complete without the mention of two pioneers in this new territory. The irony regarding each of these individuals is that they use the medium which they research as a primary vehicle for tackling this issue. Dr. Ivan Goldberg heads the on-line Internet Addiction Support Group. Goldberg has been responsible for coining the phrase “Internet Addiction Disorder,” which describes Internet addiction in clinical terms.

Psychologist Kimberly Young of the University of Pittsburgh-Bradford is leading the way for psychological evaluation of this condition and gaining the most media attention for it at this time. According to a USA Today report of July 1, 1996, Young has conducted the largest mental health study to date regarding Internet participants and Internet Addiction Disorder (IAD). The report drew the following clinical observations:

“Heavy on-line users in her study met psychiatric criteria for clinical dependence applied to alcoholics and drug addicts. They had lost control over their Net usage and couldn’t end it despite harmful effects on their personal and professional lives.” (USA Today, 1996)

Clarification of Purpose

It is important to realize that the purpose of this study was never to determine respondent addiction or dependency, as Shotton set out to do in her dependency studies. The purpose of this research has been to determine what the disruptions appear to be from those who are experiencing them the most, what factors are involved in separating the two studied groups, and then an attempt to evaluate what this might mean for a global society affected by often unlimited access to this new medium.

Respondents

Usable responses for most tests were obtained from a total of 32 confessed addicts and dependents. Data from these respondents’ self-classifications were cross-referenced with a new classification for the two primary definitions to create the independent variable used in the study. Data from all of the CAGE-related questions was added together for a median split which classified those above the median as addicts and those below the median as dependents. Respondents who answered both, not sure and neither were dropped from this portion of the study.

Data from respondents who did not fill out the CAGE sections of the McSurvey or provided incomplete data for these sections were also omitted from the final analyses of all variables in all portions of the study which used the CAGE as the independent variable.

Dependent Measures

The first dependent variable was an indexed measure of the degree of Internet use taken from five questions which addressed time on-line. Questions 1,2,3,9 and 10 looked at, on average, the number of days connected per week; on a typical day, the number of hours connected; the maximum number of hours connected on a single occasion during the previous month; times in a week connected for four or more hours on a single occasion; and, times in a month connected for four or more hours on a single occasion. A series of t-Tests was then done to see the relationship between time spent on-line by addicts and dependents.

The second dependent variable was an indexed measure of second-level CAGE components taken from questions 17-21. As with the first dependent variable, this group of questions was first considered collectively, then individually, in light of their relationship to the type of Internet user. In one instance, these data were also evaluated in light of their relationship to the respondents’ self-classifications, however with a slightly larger sample size. A series of t-Tests was done to determine the relationship between addicts and dependents regarding these particular consequences per self-classification.

Procedure

The on-line survey using human subjects was approved by the Office of Regulatory Compliance at Pennsylvania State University on June 18, 1996. The initial plan for McSurvey was its placement for public use beginning on July 7, 1996 and ending at midnight on July 28, 1996. Placement went according to the initial plan.

During the three-week period, I had made exactly 400 specific queries to individuals who should still be recognizable on-line at some sector of the World Wide Web (WWW) with terminology identifying them as addicts. The procedure for querying these individuals was simple and highly effective, though it required several repetitive macro computer techniques for expedient mastery of the entire process.

There were 120 responses to McSurvey, of which I determined 104 to be valid. Invalid responses consisted of either incomplete survey data pertinent to consistency in the overall evaluation; inability to legitimately locate a participant through the means used for survey participation; or, on a related note, out-and-out determination by an individual to slant the survey through the use and employment of systemic invalid e-mail address participation. All participants were required to submit valid e-mail response addresses, which was the only means of identifying anonymous respondents. If an address failed to show the ability to accept a simple validation response, it was considered invalid.

It is important to understand that participants answered as many questions on any portion of McSurvey which they felt like answering. Every question was given the option of no response except the e-mail validation portion which was used as a validity factor and not subject to this freedom. Therefore, many questions were left unanswered in patterns which will have to be identified at a later date in order to discern their overall importance to the finality of this research. Also, data is always poised against the total number of participants and not revealed laterally; e.g., responses of male to female were 74 and 30 respectively, actually totaling 104, but it could just as easily have been any arbitrary set of numbers not totaling 104 because respondents were free to omit data. Some questions had an option for the respondent to choose “a combination of influences” as part of an answer, thereby throwing some of those numbers into a slightly more controversial position.

Data Analysis

The first dependent variable was the indexed measure of the degree of Internet use. The questions used for this variable were designed as closely as possible to the questions which would have been used during a CAGE screen for possible substance abuse. Of the 32 respondents used in this test, 20 addicts and 14 dependents were evaluated according to their time on-line. Since the numbers were simply indexed, those who took advantage of the opportunity to reveal that they rarely or never disconnect were moved up in single increments from the highest number which occupied their respective lists; e.g., if they said they rarely disconnected on the scale for the number of days connected, they scored an eight, if they said never, they scored a nine, since seven was the highest option available for this question.

The questions which comprised the second dependent variable were indexed according to the following structure. Question 27 on socialization was simply reduced to a response of no or yes. Questions 30a and 30b were indexed to include only those responses which indicated communication skills as enhanced, crippled or neither. In other words, the not sures were dropped from responses to both of these questions, but in addition, those who responded both to 30a and 30b were dropped as well, in order to relieve their obvious redundancy.

Results

Variances by Reclassification

This sample size is 32 respondents, 20 which were deemed addicted and 14 which were deemed dependent. This group was classified as addicted or dependent by falling above or below the median on a reclassification scale which reflected their placement within the CAGE model. Crosstabs from a ChiSquare test revealed that six out of 10 self-classified addicts actually met the reclassification for addicts, while nine out of 22 dependents actually met the reclassification as dependents. This test between respondents’ self -classification, and the reclassification of addicts and dependents according to where they fell in relation to the reclassification median produced p=.96. This is further affirmation that respondents were on target in their self-classifications as addicts or dependents.

As expected, differences existed between addicts and dependents regarding the amount of time they spent on-line. A t-Test conducted using the type of Internet user classification and the time on-line as expressed in the first dependent variable revealed F(1,32)=5.8335, p< .05.

Furthermore, a t-Test using the type of Internet user classification and the amount of hours the user is on per day as expressed in Question 2 revealed F(1,32)=5.2521, p< .05.

Finally, a t-Test using the type of Internet user classification and the times per week the user is on for four or more hours as expressed in Question 9 revealed F(1,32)=6.2426, p< .05.

There were two t-Tests of the second dependent variable which found differences between the type of Internet user and consequences experienced. A t-Test between the type of Internet user classification and whether the user’s Internet connectivity has led to problems at home or school revealed F(1,32)=9.4815, p< .01.

A t-Test between the type of Internet user classification and the user’s inability to disconnect once connected produced F(1,32)=5.8401, p< .05.

Variances by Self-Classification

This sample size is 46 respondents, including 17 addicts and 29 dependents. A t-Test was done between respondents regarding their own classification as addicts or dependents, and the number of days connected per week, F(1,39)=17.3739, p< .001.

In summary, the data indicates that there are significant differences between the amount of time addicts and dependents are connected to the Internet. Furthermore, those who classify themselves as addicted or dependent hold a much stronger opinion of their actual conditions than the data warrants.

Also of significance are the consequences experienced by the two groups in relation to their time on-line. Addicts suffer more at home or school as a result of their connectivity. They also have a more difficult time getting off the Internet once connected.

The graphs indicate that addicts are actually spending less time on-line than dependents. This could presume that the time on-line may not predispose to addiction, but that other factors may be at play which are contributing to the actual addicted state.

Discussion