Measuring Television Addiction
Measuring Television Addiction
Measuring Television Addiction
Cary W. Horvath (Ph.D., Kent State University) is an Assistant Professor in the Department of Communica-
tion and Theater at Youngstown State University, t-ter research interests include interpersonal and media uses
and effects.
2004 Broadcast Education Association lournal of Broadcasting & Electronic Media 48(3), 2004, pp. 378-398
378
Horvath/MEASURINC ADDICTION 379
Addiction Defined
Television Addiction
w i t h s h knife e she took out the electrical fuses so he would stop viewing.
In Latwell, Louisiana, John Gallien shot his sister-in-law because she kept turning
down the volume" (p. 30). Studies of television deprivation (e.g.. Tan, 1977; Winick,
1988) also indicate profound and real withdrawal-like symptoms, supporting the
notion of addiction.
A h a v e have attempted to study other types of media addiction
directly using APA criteria. For example. Fisher (1994) found that children could be
vi as de o video games (based on a specific APA model of gambling
addiction). The children's pathological video game playing was based on model
criteria such a n d duration of play, supernormal expenditures, borrow-
ing and selling of possessions to play, and self-awareness of a problem. Phillips,
Rolls, Rouse, and Griffiths (1995) studied the video game habits of 868 children,
aged 11 to 16. They found that 50 could be classified as addicts. The addicted
children e v e r y every day, for longer time periods than intended, often to the
neglect of homework. They reported feeling better after play, and using play to avoid
other things (escape). Also based on APA criteria, a case study in the United Kingdom
effectively diagnosed a young man as addicted to pinball machines (Griffiths, 1992).
Consistent with third-person effect literature (e.g., Davison, 1983), the young man
thought t he o o much, but that he was not "addicted."
Therefore, a n d and inferential evidence suggests that television can be
extremely compelling and important in people's lives, even beyond dependence or
habit. Whether n viewing can truly be addictive is still unclear. Although
many have made the comparison and some have even studied addiction based on
concepts drawn from popular literature, no researchers have studied and measured
television addiction based purely on DSM-IV (American Psychiatric Association,
1994) criteria. Recently, Kubey (1996) argued that at least 5 of the 7 DSM-IV criteria
a r e t o applicable to television viewing, but this remains to be tested. Although
he did not believe th e addiction criteria of tolerance and continued use despite
e (symptoms one and seven) seemed likely for television use, he did believe
that all the others could clearly apply. According to Kubey, although we don't think
of television as a substance, we do take it into our minds. Although this is a fruitful
area of t o to diagnose television dependence have not been estab-
se (p. 233). e rn s that television use may be addictive for some people,
but addiction has not been effectively conceptualized in the communication litera-
ture (Kubey, 1996). Psychiatry has provided criteria for dependence/addiction that
have taken decades to develop, but communication scholars have yet to attempt to
use them fully.
In that effort, this research involved two separate studies. The first used conve-
nience sampling to pilot and test the validity of a new measure of television
addiction. The second employed purposive sampling, replicated parts of the first
study, and added a measure of social desirability. Reliable multidimensional and
unidimensional scales f television addiction result from the research.
Horvath/MEASURINC ADDICTION 383
Study One
Participants
content gleaned from popular literature, not psychiatry. Here, five statements were
created o represent each f the seven criteria of addiction/dependence: tolerance
(e.g., "I feel like I watch more TV than I used to in order to feel the same");
withdrawal (e.g., "when I am unable to watch television, I miss it so much that you
could call it 'withdrawal'"); unintended use (e.g., "I often watch TV for a longer time
than I intended"); cutting down (e.g., "I often think that I should cut down on the
amount of television that I watch"); time spent (e.g., "compared to most people, I
spend a great deal of time watching television"); displacement of other activities
(e.g., "I often watch television rather than spending time with friends and family");
and continued use (e.g., "I keep watching TV even though it is causing serious
problems in my life"). Response options were strongly agree (5), agree (4), agree
some and disagree some (3), disagree (2), and strongly disagree (1). The 35-item
measure appeared to have face validity, as assessed by the researcher and a
colleague in communication studies (see Appendix A for the measure).
Principal components factor analysis with iterations and varimax rotation deter
mined the factor structure for the addiction measure. (Oblimin rotation yielded no
improvement from the resultant varimax structure.) A factor needed a minimum
eigenvalue of 1.0 and t least three loadings meeting a 50/40 rule (in which an item
loads on o r at .50 or more, and less than .40 on other factors) to be retained.
A liberal 50/40 rule was applied because the resultant factors approximated DSM-IV
(American Psychiatric Association, 1994) criteria, and to retain as many of the
original items as possible. The analysis initially identified six factors that explained
62.59% of the total variance. However, only four factors explaining 55.45% of the
variance were retained for analysis. Retained factor items were summed and aver-
aged to create factor indices. The primary loadings in the factor solutions are
summarized in Table 1.
Factor 1, Problem Viewing, accounted for 22.1% of the total variance after
rotation (eigenvalue = 7.75). Its 11 loadings described television viewing that is
problematic (e.g., has created real problems for me, family members get angry; M
= 1.64, SO = 0.58, Cronbach a = .93). Of the seven DSM-IV criteria for
psychological addiction (American Psychiatric Association, 1994) all five of the
items designed to represent "continued television use despite problems" loaded on
this factor. Two items from each of the following criteria also loaded on this
factor: time spent, cutting down, and displacement of other activities.
Factor 2, Heavy Viewing, accounted for 14.9% of the total variance after rotation
(eigenvalue = 5.20). Unlike Factor 1, these 10 items represented heavy viewing, but
not necessarily m viewing (e.g., time really gets away from me, wind up
spending hours; M = 2.41, SD = 0.78, a = .89). The APA criteria were represented
in this factor by three displacement items, three unintended use items, two time spent
items, and two cutting down items.
Factor 3, Craving for Viewing, accounted for 7.8% of the total variance after
rotation (eigenvalue = 2.74). The six items that loaded on this factor suggested a
craving to watch more television (e.g., I've tried to reduce, I watch more and more
Horvath/MEASURINC ADDICTION 385
Item Number 1 2 3 4
shows lately; M = 2.00, SD = 0.60, a = .75). The APA criteria for this factor
included four tolerance items (one tolerance item was inadvertently repeated in the
measure), one cutting down item, and one withdrawal item.
Factor 4, Withdrawal, accounted for 7.3% of the total variance after rotation
(eigenvalue = 2.57). The four items loading on this factor denoted feelings of
withdrawal when without television (e.g., I miss it so much, can't imagine going
without TV; M = 2.36, SD = 0.81, a = .65). The APA criteria items loading on this
factor were three withdrawal items and one time spent item.
Factor 5 was not retained due to its low reliability (a = .53). Also, a one-item
Factor 6 was not d because it failed to meet the specified criteria.
Item analysis was used o improve upon the reliability of factor structures. A few
changes were made to the initial structure. For example, statement 1 ("I feel like 1
watch more TV than I used to in order to feel the same enjoyment"), which loaded
on Factors 1 and 3, was placed in Factor 3 (see Table 1 for the factor structure and
Appendix A for I t s Its removal from Factor 1 reduced the reliability coefficient
from .95 to .94, but its addition to Factor 3 increased the reliability coefficient from .
70 to .75. Also, the 50/40 rule was relaxed to include items that made conceptual
sense. For m item 6, "When I am unable to watch TV, I miss it so much that
you could call it 'withdrawal,'" did not reach the .50 mark but most clearly loaded
on Factor 4 with other withdrawal items. Similarly, items 21 and 25 loaded on
Factors 1 and 2, but because they were time spent items that positively contributed
to reliability, they were retained on Factor 2.
As a result of the pilot study, a few items that failed to fit the factor structure were
reworded for clarity, as suggested by Kerlinger (1986). Item 2, "\ watch about the
same amount of TV as I used to" was changed to "I've watched the same amount of
TV as I always have." Item 8, "If I have to miss my shows for some reason, it makes
me very upset" became "If I have to miss a favorite show I feel upset," and item 15,
"I usually a TV for s long as I planned" was extended to "I usually watch TV
for t o as I planned to watch." In addition, a few of the items that loaded on
Factor 1 were altered to include the word "sometimes," with the hope of increasing
the factor variance. Appendix A reflects these revisions.
morning?). Two or three affirmative responses to the measure indicates high suspi-
cion of alcoholism. According to Ewing (1984), "even one positive reply calls for
further inquiry" (p. 1907). Responses were coded 1 = no and 2 = yes.
h a s has been widely used because of its ease of
administra-
tion and effectiveness (see Ewing, 1984) and has been validated by others (e.g.,
Mayfield, McLeod, & Hall, 1974). The measure has been found to be more sensitive
than other screening devices, with accuracy ratings in predicting alcoholism of over
90% (Ewing, 1984). In s study, 28% of the participants answered "yes" to question
1, % to to 2, 11 % to question 3, and 28% to question 4. Because high scores
on CAGE and high scores on the Television Addiction Measure suggest television
addiction, they should be positively correlated.
Results
p < .001), Craving for Viewing (r = .29, p < .001), and Withdrawal (r = .50, p <
.001). wa s was also positively related to CAGE scores (r = .28, p <
.001). So, those who watched more hours of television were very likely to feel
withdrawal without it, and likely to report heavy viewing, problem viewing, craving
for viewing, d respond affirmatively to CAGE items.
Also, relationships emerged between Withdrawal and usually watching in the
evening (r = .42, p < .001), as well as yesterday evening (r = .37, p < .001).
Problem Viewing was also related to time spent viewing yesterday evening {r = .34,
p < .001) and usually in the evening (r = .31, p < .001). In general, correlations for
time t he in the evening were higher for all television addiction factors.
t h i s from this preliminary analysis, people who watch heavily in the
evening hours tend to feel withdrawal from television more when it's gone and
experience m problem viewing than other people.
Statistically significant, but weak correlations resulted between the Television
Addiction Scale factors and demographic variables. For example, age, in number of
years, was positively related to Craving for Viewing (r = .18, p < .001), and gender
was negatively related to Problem Viewing (r = -.17, p< .01). So, there was a slight
e f or t o t o crave viewing more than younger people and for males
to experience problem viewing more than females. Also, education was weakly and
negatively related to Problem Viewing (r = .13, p < .05) and Craving for Viewing
(r = -14, p < .05). Therefore, a slight tendency resulted for more educated people
t o r e p o r t f r o m from viewing or a craving for viewing. In comparing age,
total E and CAGE items, average exposure was positively related to age (r = .
24, p < .001) and to CAGE total scores (r = .28, p < .001). However, age was
unrelated to CAGE total scores (r = .03, p = .58). Therefore, people who watched
heavy amounts of television tended to be older, and answer affirmatively to CAGE
thos e often wh o who watch light amounts of television.
Two-tailed f-tests were used to compare scores by gender on the two measures of
television addiction. Scores on the CAGE instrument (Ewing & Rouse, 1970) were
independent of gender effects, t{292) = 1.57, p = .12. Similarly, three of the
Television Addiction Scale factors were independent of gender effects, including
Heavy Viewing, ((296) = 1.54, p= .12, Craving for Viewing, ((296) = 1.81, p= .07,
and Withdrawal, #296) = 1.24, p = .22. On Factor 1, Problem Viewing, scores did
differ by gender, #296) = 3.01, p = < .01. Specifically, men {M = 1.70) scored
somewhat higher than women (M = 1.50) on problem viewing.
Study Two
Although the first study indicated that a valid and reliable measure of television
addiction could result from convenience sampling, it was unable to address the
stability of findings or how social desirability might have affected results. A second
Horvath/MEASURING ADDICTION 389
study was conducted to sample more purposefully, to replicate parts of the first study,
reduce error, a to add a measure of social desirability.
Participants
Measurement
The same adapted version of the CAGE questionnaire (Ewing & Rouse, 1970) was
again indexed (M = 5.14, SD = 1.17). Last, the Marlowe-Crowne Social Desirability
Scale Form C C Form ) (Reynolds, 1982) was added. Although other short forms
of the original 33-item Marlowe-Crowne Social Desirability Scale (Crowne & Mar-
lowe, 1960) s e e Strahan & Gerbasi, 1972), Reynolds found Form C to be
best, based on comparison of seven different versions of the measure. A 13-item
measure, it requires participants to respond to statements such as "I am always
courteous, even to people who are disagreeable" and "There have been occasions
when I o advantage f someone." The usual response method is true-false.
However, this study employed a response format consistent with the other measures
used. Therefore, response options included strongly agree (5), agree (4), disagree
some and agree some (3), disagree (2), and strongly disagree (1).
Using Kuder-Richardson formula 20 reliability, Reynolds (1982) found that the
M-C Form C reliability was acceptable (.76). The measure was also shown to be
highly related to the original Marlowe-Crowne Standard (r = .93, p < .001) and
related to the Edwards Social Desirability Scale (Edwards, 1957) in a manner
consistent with the original (r = .41, p < .001). In this study, the instrument's
t w ea k (Cronbach a = .59), so item analysis was used to
eliminate item 1, "It is sometimes hard for me to go on with my work if I am not
encouraged," and item 3, "On a few occasions, 1 have given up doing something
because I thought too little of my ability." The resultant 11-item measure was more
reliable (M = 3.19, SD = 0.53, a = .74).
Television addiction was measured by the revised version of the Television
Addiction Scale (see Appendix A). Items were reordered before being included in the
questionnaire. To replicate the pilot study, principal components factor analysis with
iterations and varimax rotation determined the factor structure for the addiction
measure (oblimin rotation yielded no improvement from the resultant varimax
structure), e more conservative rules for factor analysis guided this study so
that the most efficient measure of television addiction would emerge. This time, a
factor needed a m eigenvalue of 1.0 and at least three loadings meeting a
60/40 rule to be retained. The 60/40 rule was relaxed for a few items that approx-
imated the rule, made conceptual sense, and added to factor reliability. The analysis
identified six t that explained 64.6% of the total variance. However, only two
of the factors explaining 49.98% of the variance were retained for analysis. Retained
factor items were summed and averaged to create factor scores. See Table 2 for a
summary of primary factor loadings.
Factor 1, Problem Viewing, accounted for 24.6% of the total variance after
rotation (eigenvalue = 8.61). Its 14 loadings described television viewing that is
Horvath/MEASURING ADDICTION 391
.82 .17
.81 .26
.79 .29
.79 .26
.72 .30
.71 .24
.70 .33
.65 .00
.64 .20
.64 .32
.61 .14
.61 .30
.60 .39
.54 .52
.40 .71
.34 .68
.37 .61
.37 .58
.46 .57
.30 .56
Factors
Item Number
Factor 1/Problem Viewing
31 causing serious problems
32 family members get angry
35 loved ones can't stand it
34 created real problems
03 watch more and more
28 whole life revolves around
33 alienating my loved ones
33 you could call it withdrawal
34 watch because I missed it
19 feel bad but can't stop
01 more and more to feel same
25 more time than anything
35 embarrassed to tell people
36 great deal of my time
Factor 2/Guilty Viewing
29 more productive if I didn't watch
16 think I should cut down
13 wind up spending hours
18 guilty about watching so much
27 should be working/school
30 would spend time with hobbies
Item Number 1 2
Eigenvalues after oblimin rotation 10.22 8.70
TV Addition items
06 you could call it withdrawal
19 feel bad but can't'stop
07 watch because I missed it
20 embarrassed to tell people
33 alienating my loved ones
01 watch more t o to feel the same
34 has created real problems for me
25 more time than anything else
32 family members get angry 03
watch more and more lately
35 loved ones can't stand it
28 whole life revolves around the TV
31 causing serious problems in my life
21 great deal of time
30 would spend more time with hobbies
18 guilty about watching so much 16
think I should cut down
29 more productive if I didn't watch
27 should be working/going to school
13 wind up spending hours
72 .45
0 .70
76 .59
7 .67
76 .62
72 .48
83 .61
70 .62
84 .57
75 .61
8 .58
81 .60
5 .55
67 .74
49 .73
58 .77
54 .78
56 .82
62 .73
59 .80
Horvath/MEASURINC ADDICTION 393
Results
The 20-item Television Addiction Scale was positively related to Ewing and
Rouse's (1970) CAGE instrument (r = .57, p < .001) and negatively related to the
measure of social desirability (r = -.27, p < .001). Interestingly, older participants re
por likely t report socially desirable answers (r = .21, p < .001). As
expected, participants who were identified as "heavy" (coded 2) or "light" (coded 1)
viewers by student research assistants differed on measures of television addiction.
Independent f-tests showed that people classified as heavy viewers (M = 2.22)
reported more television addiction than light viewers (M = 1.66), H297) = 7.16,
p < .001, and had higher scores {M = 5.64) on the CAGE instrument than light
viewers (M = 4.65), f(315) = -8.67, p < .001. Although exposure was related to age
(r = .17, p < .01), Pearson correlations indicated that no relationships existed
between television addiction and age (r = .03, p = .61) or television addiction and
education (r = .07, p = .22). Results from an independent f-test indicated that men
(/Vl = 2.07) scored significantly higher than women (M = 1.82) on the television
addiction measure, f(31) = 2.88, p < .01. Men {M = 5.32) also scored higher than
women (M = 5.00) on CAGE scores, f(308) = 2.51, p < .05. However, mean
differences were rather small. So, gender produced the only demographic difference
in television addiction scores, as measured by the Television Addiction Scale and the
CAGE instrument.
Television exposure was positively related to television addiction (r = .47, p <
.001) and ( r = .37, p < .001). Einally, a partial correlation between
Te l e v i s i o n d and the CAGE measure, controlling for gender and social
t revealed a a strong relationship still existed (r = .54, p < .
001).
Discussion
as heavy viewers reported more television addiction and had higher scores on the
screening scale than light viewers. No relationships emerged between the 20-item
e an scale d age or education; however, men scored significantly
higher than women.
One problem i n this research was the potential for participant bias
toward researchers for some variables. Eor example, some participants made verbal
remarks th e y assumed the researcher was against television viewing.
Taken together with the negative relationship between social desirability and tele-
vision addiction, it is possible that Type II error resulted in this study; more robust
effects may exist, but were not identified by this analysis because of minimal
variance on the measure. To minimize the problem of socially desirable responses,
future studies should continue to use established scales to measure and control for
the effects of social desirability. It might also help to provide a very clear explanation
of the purpose of research, and encourage people to respond honestly about their
television use. For example, instructions could be more explicit in cover letters,
clearly explaining th e researcher does not necessarily have a personal bias
against media. Perhaps a cover letter that spoke somewhat favorably about television
use would discourage socially desirable answers. Such statements might range from
obvious to subtle in nature.
Also, because we know very little about how television addiction might manifest
in the child or adolescent population, research with those samples should yield
interesting results. It seems likely that adolescents are capable of being addicted to
television. Typologies f viewing patterns have been created on demographic bases,
finding h a s as homemakers, children, and elderly people are more
dependent on television (Donohew, Palmgreen, & Rayburn, 1987; Frank & Green-
berg, 1980). Rubin (1977) found that in the adolescent population, habit and pass
time are major viewing motivators. Media scholars need to determine children's
susceptibility to psychological addiction to television at different developmental
stages in life, whether children's programming can lead to or encourage addiction,
and whether children, if addicted, experience greater media effects such as purchas
ing advertised goods, attitude formation, or social learning from television. We need
to discover possible detrimental effects children and adolescents might experience in
school, at home, and at work. We also need to establish what adults can do to reduce
addictive a s as media literacy projects or parental intervention. Euture
research might also adapt the Television Addiction Scale to other and more specific
kinds of media addiction such as sports, soap operas, romance novels, newspapers,
radio, or the e The measure of television addiction is flexible enough to adapt,
while reflecting the DSM-IV criteria (American Psychiatric Association, 1994).
This a continues a process of teasing apart the concepts of normal, heavy,
and problem uses of television that confound the media literature. There remains
much work to be done in this area in order to determine whether television addiction
may be a s "true" addiction, such that people manifest legitimate
Horvath/MEASURING ADDICTION 395
Appendix Television
Addiction Scale
Tolerance
37 I feel like I watch more TV than I used to in order to feel the same.**
38 I've watched the same amount of TV as I always have. ***| rve watched the same amount of TV
as I
used to.]*
39 It seems like I watch more and more shows lately for the same amount of enjoyment.
40 I watch amoun t same o f TV as I used to but 1 don't get the same effect from it
anymore.
41 I watch more and more shows to try to feel the same as I used
to.
Withdrawal
42 When I am unable to watch television, I miss it so much that you could call it "withdrawal."
43 Sometimes I watch TV just because I missed it a great deal.
44 If I have to miss a favorite show, I feel upset. [If I have to miss my shows for some reason, it
makes
me very upset.]
45 I can't imagine going without TV.
10. I could easily go without TV with no problem.*
Unintended Use
11.1 often watch TV for a longer time than I intended.
46 Time really gets away from me when I watch TV.
47 Sometimes I only plan to watch TV for a few minutes, and wind up spending hours in front of it.
48 I follow a very exact TV-watching schedule.*
49 I usually watch TV for exactly as long as I planned to watch. |l usually watch TV for as long as I
planned.]*
396 Journal of Broadcasting & Electronic Media/September 2004
Appendix (Continued)
Cutting Down
50 I often think that I should cut down on the amount of television that I watch.
51 I've tried to reduce the amount of TV I watch, but it hasn't really worked.
52 I often feel guilty about watching so much television.
53 I feel bad that I watch so much TV, but I can't seem to stop.
54 I would be embarrassed to tell people how much TV I actually watch.
Time Spent
55 Compared to most people, I spend a great deal of time watching television.
56 Much of my time is spent in front of the television.
57 Television viewing takes up almost all of my leisure time.
58 I don't spend that much time watching television. *
59 I spend more time watching TV than just about anything else.
Displacement of Other Activities
60 I sometimes watch television when I should be spending time with friends or family.
61 I often watch TV when I should be working or going to school.
62 Sometimes I feel like my whole life revolves around the TV, and I never do anything else.
63 I would be a lot more productive if I didn't watch so much TV.
64 I would spend more time with hobbies if I didn't watch so much TV.
Continued Use
31.1 keep watching TV even though it is causing serious problems in my life.
65 My family members get angry and tell me I watch too much TV, but I can't stop.
66 I sometimes feel like my TV watching is alienating my loved ones.
67 My TV watching has created real problems for me, but i keep watching.
68 I keep watching TV even though my loved ones can't stand it.
Note: *ltems 2 10, 4 15, and 24 are reverse-scored. **ltems shown in italics are retained for the 20-item
TV Addiction * ** Bracketed phrases reflect wording from pilot study.
References
Akers, R. L. (1991). Addiction: The troublesome concept. Journal of Drug Issues, 21, 777-793.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders
(4th ed.). Washington, DC: Author. Anderson, . R., A. , A., Schmitt, K. L., & Jacobvitz,
R. S. (1996). Stressful life events and
television viewing. Communication Research, 23, 243-260. Beresford, T. (1991). The
nosology of alcoholism research. Alcohol Health and Research
World, 15, 260-271. Bingham, J. E., & Piotrowski, C. (1996). On-line sexual addiction:
A contemporary enigma.
Psychological Reports, 79, 257-258.
Condry, J. (1989). The psychology of television. Hillsdale, NJ: Eribaum. Crowne, D. P., &
Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. journal of Consulting Psychology, 24, 349-354. Davison, W. P.
(1983). e third-person effect in communication. Public Opinion Quarterly,
47, 1-15. DeFleur, M. L, & Ball-Rokeach, S. J. (1989). Theories of mass communication
(5th ed.). New
York: Longman. Donohew, L, P. , & Rayburn, J. D., II. (1987). Social and
psychological origins of
media use: A lifestyle analysis. Journal of Broadcasting & Electronic Media, 31, 255-278.
Edgar, P. (1977). Families without television. Journal of Communication, 27, 73-77. Edwards,
A. (1957). The social desirability variable in personality assessment and research.
New York: Dryden. Elliott, R., Eccles, S., & Gournay, K. (1996). Revenge,
existential choice, and addictive
consumption. Psychology & Marketing, 13, 753-768. Ewing, J. A. (1984). Detecting
alcoholism: The CAGE questionnaire. Journal of the American
Medical Association, 252, 1905-1907.
Horvath/MEASURING ADDICTION 397
Ewing, J. A., & Rouse, B. A. (1970, February). Identifying the hidden alcoholic. Paper presented
at the o n on Alcohol and Drug Dependence, Sydney, Australia.
Faber, R. J., C , C, & Krych, R. (1987). Compulsive consumption. Advances in
Consumer Research, 79,459-469.
Fisher, S. (1994). g video game addiction in children and adolescents. Addictive
Behaviors, 19, 545-553.
Foss, K. . & F . F. (1996). Exploring the margins of television viewing. Commu-
nication Reports, 9, 61-68. Fowles, J. (1982). Television viewers vs. media snobs: What TV
does for people. New York:
Stein and Day.
F r a n k , M . & G . G. (1980). The public's use of television: Who watches and why.
Beverly Hills, CA: Sage. Griffiths, M. D. (1992). Pinball wizard: The case of a pinball
machine addict. Psychological
Reports, 71, 160-162.
Horton, D., & Wohl, R. R. (1956). Mass communication and para-social interaction: Observa-
tions on intimacy at a distance. Psychiatry, 19, 215-229. Keller, M., & Doria, J. (1991). On
defining alcoholism. A/co/io/ Hea/f/i and Research World, 15,
253-259. Kerlinger, F. N. (1986). Foundations of behavioral research (3rd ed.). Fort
Worth, TX: Holt,
Rinehart and Winston. Kubey, R. W. (1996). Television dependence, diagnosis, and
prevention. In T. M. MacBeth
(Ed.), Tuning in to young viewers (pp. 221-260). Thousand Oaks, CA: Sage. Kubey, R., &
Csikszentmihaiyi, M. (1990). Television and the quality of life: How viewing
shapes everyday experience. Hillsdale, NJ: Eribaum. Kubey, R. W., J. , J., & Barrows,
J. R. (2001). Internet use and collegiate academic
performance decrements: Early findings. Journal of Communication, 57(2), 366-382.
Mander, J. (1978). Four arguments for the elimination of television. New York: William
Morrow.
Mayfield, D. G. , G., & Hall, P. (1974). The CAGE questionnaire: Validation of a new
alcoholism screening instrument. American Journal of i'sychiatry, 131, 1121-1123.
Mcllwraith, R. D. (1998). "I'm addicted to television": The personality, imagination, and TV
watching patterns of self-identified TV addicts. Journal of Broadcasting & Electronic Media,
42, 371-386. Mcllwraith, R . R. S., Kubey, R., & Alexander, A. (1991).
Television addiction:
Theories and t h e ubiquitous metaphor. American Behavioral Scientist, 35,
104-121.
Morgan, M., & N. N. (1990). Cultivation analysis: Conceptualization and methodol -
ogy. In N. Signorielli & M. Morgan (Eds.), Cultivation analysis: New directions in media
effects research (pp. 13-34). Newbury Park, CA: Sage.
Morris, K. (1998). Seeking ways to crack cocaine addiction. Lancet, 352, 1290. Peele, S.
(1985). The meaning of addiction. Lexington, MA: Lexington Books. Phillips, C. A., Rous e
, A. ,c & Griffiths, M. D. (1995). Home video game playing in
school children: A study f incidence and patterns of play. Journal of Adolescence, 18,
687-691.
Rasmussen, S. (2000). Addiction treatment: Theory and practice. Thousand Oaks, CA: Sage.
Reynolds, W. M. (1982). Development of reliable and valid short forms of the Marlowe-Crowne
Social Desirability Scale. Journal of Clinical Psychology, 38, 119-125. Rubin, A. M.
(1977). Television usage, attitudes and viewing behaviors of children and
adolescents. Journal of Broadcasting, 21, 355-369. Rubin, A. M. (1981). An
examination of television viewing motivations. Communication
Research, 8, 141-165. Rubin, R. B., Rubin, , M., & Piele, L. J. (1990). Communication
research: Strategies and
sources (2nd ed.). Belmont, CA: Wadsworth.
398 Journal of Broadcasting & Electronic Media/September 2004