Three hundred and fifty three persons responded to the online questionnaire posted on the Internet.  These persons were self selected individuals who agreed to a consent form before participating in the questionnaire.

            The age of the participants ranged from 14 to 72 years.  The mean age was 27.84 (SD = 11.52).  Approximately seventy percent of the sample was within ages 14 to 30.  One possible reason for this uneven distribution is because the Internet is a medium more extensively used by younger individuals.  For the purposes of this research, the adolescent period was defined as the period from ages 14-20 (34.3%), young adulthood ages 21-29 (34%) and adulthood from ages 30 to 72 (31.7%).  There were 111 males (31.4%) and 242 females (68.6%).

            Overall, participants from thirty-one countries took part in this research (see Table 2).  However, the spread of participants from the different countries was unevenly distributed.  Seventeen of the 31 countries had only one participant and 72.2% of the sample came from the United States.  Because such a large proportion of the sample came from one particular country, it was difficult to create any meaningful or statistically useful groups of nationality.  Therefore nations were grouped into two major categories, one group representative of an industrialized, Western and individualistic society and the other group representative of countries that are either less industrialized or have more of a collectivist culture.  The researcher decided a country’s classification into one or the other category.  The final sample of nations was separated with 87.6% falling into the first category, industrialized societies and 12.4% falling into the latter category, collectivistic societies.

            The ratio of males to females was approximately the same for the two categories of countries, however, age distribution was dissimilar from industrialized societies to collectivist societies.  For the industrialized societies category, the percentage of males was 31% and in collectivistic societies it was 34%.  With regards to age, participants in the industrialized societies group showed a relatively even distribution among the three age groups (37%, 29%, 34%) while collectivist societies showed a more uneven distribution with the majority falling within the young adult age category (16%, 66%, 18%).

Table 2

Frequency of Nationality of Respondents

 Procedure

            An online questionnaire was used to gather data about Internet use, coping strategies, the level of loneliness, and demographics.  The questionnaire was part of a larger website dedicated to providing information about loneliness.  This larger website formed part of the recruiting process for participants.  The questionnaire was open to anyone with Internet access age 14 and over who had agreed with the consent form.  Recruitment of participants and advertisement of the questionnaire took place in several ways:

1.      Advertisements were made on research websites, that is, websites that posted online research projects.

2.      Advertisements were also made on websites that are affiliated with loneliness in some way, for example, websites that are dedicated to self-help, or relationship problems.

3.      The website was also advertised through search engines.  The website address and a short description were submitted to several search engines.  Visitors then found the website through searches performed on these search engines especially searches that included the word “loneliness” in it.

4.      An attempt was made to recruit persons who had personal websites that in some way dealt with loneliness.  This was usually in the form of a webpage that had a poem or other artwork about loneliness.  Creators of these websites received an email inviting them to visit the website and share their loneliness experiences.

5.      Recruitment was also made through online support groups, in particular, newsgroups and clubs in Yahoo.  Members either received an email or read a message that was posted on the club message board, inviting them to visit the website and share their loneliness experiences.

6.      Advertisements made through recommendations made by persons who visited the site.

7.      Advertisement to students at the University of Illinois in classes they attended.

8.      Advertisement to friends.

An Internet counter, located on the introductory page, provided partial information on the rates of visitors from referring URLs.  This information suggested that advertisement procedures 1, 2 and 3 above were the most popular ways participants found the site.  Advertisements made directly to persons brought vast increases in participation but was only temporary and faded after time.

The online questionnaire was constructed so that respondents would only be able to submit the questionnaire when all the necessary questions were answered.  After the questionnaire was answered respondents were invited to purview some preliminary results of the data collected.  They were also provided with a unique random questionnaire number.  Respondents who decided to answer the questionnaire more than once were asked, in the questionnaire, if they answered the questionnaire before and to provide, if possible, the previous questionnaire number they were assigned.  This was one technique used to portion out repeated participation by the same person.  Five persons reported answering the questionnaire more than once, with responses the second time being almost identical to the first.  The first questionnaire responses were deleted before data analysis commenced.

Measures

      Variables of interest were: the strength of loneliness, types and strengths of coping strategies used, Internet use, and demographics.

      Strength of loneliness.  The strength of loneliness was measured using an adapted version of the Revised UCLA Loneliness Scale (Russell, Peplau & Cutrona, 1980) plus one extra item.  The Revised UCLA Loneliness scale measures loneliness by asking indirect questions about experiences associated with loneliness.  The extra item asked directly whether respondents feel lonely or not.  The final instrument consisted of eleven items.  Responses to each item ranged from often feel this way to never feel this way on a four point Likert scale.  Wilson, Cutts, Lees, Mapungwana and Maunganidze (1992) reported good reliability with eight of these eleven items.  The scale itself is one of the most widely used loneliness measures and has reputable reliability and validity (Shaver & Brennan, 1991).

      The loneliness scale without the extra item had an acceptable internal consistency, α = .92.  The single item that measured loneliness by directly asking participants if they were lonely had a corrected item-total correlation of .77 with the rest of the scale and raised the internal consistency of the final measure to .93. 

Types and strengths of coping strategies.  The sixty item instrument measured 28 different coping strategies (see Appendix A).  These 28 coping measures represented both coping strategies used in previous loneliness research as well as other strategies used in non-loneliness research such as daydreaming, obtaining power and revenge, and venting one’s emotions.  The final instrument was a combination of three different coping instruments along with one additional item to measure the Internet as a coping strategy.  The three different coping instruments were the Sustaining Fantasy Instrument (Zelin et al., 1983), the COPE scale (Carver, Scheier & Weintraub, 1989) and the Reactions To Loneliness Measure (Rubenstein & Shaver, 1982).  The combined instrument was designed to fulfill Objective 1. 

Of the 28 coping strategies, 13 were single item questions.  Most of these single items were coping strategies used in the study by Rubenstein and Shaver (1982).  The remaining 15 measures all had good internal reliability with adjustments made to two measures.  Reliabilities ranged from .68 to .98 (see Table 3).

Table 3

Reliabilities of 15 coping strategies

 Note. ^* One item was dropped from scale to increase reliability.

Internet Use.  Researchers have created several ad hoc measures of Internet use.  For this research, the categories of Internet use were derived from several informal Internet use measures.  Two dimensions of Internet use were of interest:  different ways participants use the Internet and what they considered to be the most important use of the Internet for them.  To measure these two dimensions, two different questions were asked.  The first question (Appendix A, question 7) allowed participants to select as many uses (up to 12) of the Internet that were applicable to them.  The second question (Appendix A, question 8) asked participants about the most important use of the Internet to them.  For this second question they were presented with the same 12 items in the previous question as possible choices but were only allowed to select one.