Aguilar, R. (1996). Net’s sleazy rap persists. [Online]. News.com, November 1.

Barge, J. K., & Musambira, G. W. (1992). Turning points in chair -faculty relationships. Journal of Applied Communication Research, 20, 54-77.

Baxter, L. A., & Bullis, C. (1986). Turning points in developing romantic relationships. Human Communication Research, 12, 469-493.

Baxter, L. A., & Erbert, L. A. (1999). Perceptions of dialectical contradictions in turning points of development in heterosexual romantic relationships. Journal of Social and Personal Relationships, 16, 547-569.

Bell, R. A., & Daly, J. A. (1984). The affinity -seeking function of communication. Communication Monographs, 51, 91-115.

Benedetti, W. (2000). Online dating presents a new set of potential dangers for the unwary. Seattle Post Intelligencer, February 7.

Berscheid, E. (1985). Interpersonal attraction. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology (pp.413-484). New York: Random House.

Bolton, C. D. (1961). Mate selection as the development of a relationship. Marriage and Family Living, 23, 234-240.

Brehm, S. (1992). Intimate relationships. New York: McGraw-Hill.

Bruess, C. J. S., & Pearson, J. C. (1997). Interpersonal rituals in marriage and adult friendships. Communication Monographs, 61, 25-46.

Bullis, C., & Bach, B. W. (1989). Are mentor relationships helping organizations? An exploration of developing mentee-mentor-organizational identifications using turning point analysis. Communication Quarterly, 37, 199-213.

Bullis, C., Clark, C., & Sline, R. (1993). From passion to commitment: Turning points in romantic relationships. In P. J. Kalbfleisch (Ed.), Interpersonal communication: Evolving interpersonal relationships (pp. 213-237). Hillsdale, NJ: Laurence Erlbaum.

Bulmer, M. (1979). Concepts in the analysis of qualitative data. Sociological Review, 27, 651-677.

Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-38.

Cooper, A., & Sportolari, L. (1997). Romance in Cyberspace: Understanding online attraction. Journal of Sex Education and Therapy, 22, 7-14.

Culnan, M. J., & Markus, M. L. (1987). Information technologies. In F. M. Jablin, L. L. Putnam, K. H. Roberts, & L. W. Porter (Eds.), Handbook of organizational communication: An interdisciplinary perspective (pp. 420-443). Newbury Park, CA: Sage.

Daft, R. L., & Lengel, R. H. (1986). Information richness: A new approach to managerial behavio and organization design. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior (vol. 6, pp. 191-233). Greenwich, CT: JAI.

Dubrovsky, V. J., Kiesler, S., & Sethna, B. N. (1991). The equalization phenomenon: Status effects in computer-mediated and face-to-face decision-making groups. Human Computer Interaction, 6, 119-146.

Finholt, T., & Sproull, L. S. (1990). Electronic groups at work. Organization Science, 1, 41-64. French woman dies after an Internet romance sours. (1999, March 15). Chicago Tribune.

Gerhardt, P. (1999). Sex, lies & e-mail: Internet dating offers plenty of opportunities to meet that special someone, but therapists warn that what you see online is not always what you get. The Washington Post, July 27.

Granovetter, M. S. (1976). Network sampling: Some first steps. American Journal of Sociology, 81, 1287-1303.

Hahn, H., & Stout, R. (1994). The Internet complete reference. Berkeley, CA: Osborne McGraw-Hill.

Hamilton, J. O. (1999). When cupid uses a cursor. Business Week, February 22.

Hatfield, E., & Sprecher, S. (1986). Mirror, mirror… The importance of looks in everyday life. Albany: SUNY Press.

Heim, M. (1992). The erotic ontology of cyberspace. In M. Benedikt (Ed.), Cyberspace: First steps (pp. 59-80). Cambridge, MA: MIT Press.

Hendrick, C., & Hendrick, S. (1983). Liking, loving and relating. Monterey: Brooks/Cole.

Internet romance ends with man jailed in Wales. (1999, June 19). Seattle Times, p. A11.

Huston, T. L., Surra, C., Fitzgerald, N. M., & Cate, R. (1981). From courtship to marriage: Mate selection as an interpersonal process. In S. Duck & R. Gilmour (Eds.), Personal relationships 2: Developing personal relationships. New York: Academic Press.

Jacobson, D. (1996). Contexts and cues in cyberspace: The pragmatics of naming in text-based virtual realities. Journal of Anthropological Research, 52, 461-479.

Kiesler, S. (1986). The hidden messages in computer networks. Harvard Business Review, 64, 46-54, 58-60.

Lloyd, S. (1983). A typological description of premarital relationship dissolution. Unpublished doctoral thesis, Oregon State University.

McCormick, N. B., & McCormick, J. W. (1992). Computer friends and foes: Content of undergraduates’ electronic mail. Computers in Human Behavior, 8, 379-405.

McGrath, J. E. (1991). Time, interaction, and performance (TIP): A theory of groups. Small Group Research, 22, 147-174.

Montgomery, C. (1999). Love blooms in cyberland. Insight on the News, February 8.

Morris, B. R. (1999). You’ve got romance! Seeking love on line. New York Times, August 26.

Negroponte, N. (1995). Being digital. New York: Alfred A. Knopf. Parks, M. R. (1997). Communication networks and relationship life cycles. In S. Duck (Ed.), Handbook of personal relationships (2 nd ed.). London: John Wiley.

Parks, M. R., & Floyd, K. (1996). Making friends in cyberspace. Journal of Computer-Mediated Communication, 1(4). [On-line]. Available: http://jcmc.huji.ac.il/vol1/issue4/vol1no4.html

Parks, M. R., & Roberts, L. D. (1998). “Making MOOsic”: The development of personal relationships on line and a comparison to their offline counterparts. Journal of Social and Personal Relationships, 15(4), 517-537.

Rheingold, H. (1993). The virtual community: Homesteading on the electronic frontier. Reading, MA: Addison-Wesley.

Schlosberg, S. (1999). For a good man, click here. Health, September.

Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. London: Wiley.

Snyder, M., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self-fulfilling nat ure of social stereotypes. Journal of Personality and Social Psychology, 35, 656-666.

Sproull, L., & Kiesler, S. (1986). Reducing social context cues: Electronic mail in organizational communication. Management Science, 32, 1492-1512.

Surra, C., Arizzi, P., & Asmussen, L. (1988). The association between reasons for commitment and the development and outcome of marital relationships. Journal of Social and Personal Relationships, 5, 47-63.

Turkle, S. (1995). Life on the screen: Identity in the age of the Internet. New York: Simon & Schuster.

Walther, J. B. (1992). Interpersonal effects in computer-mediated interaction: A relational perspective. Communication Research, 19, 52-90.

Walther, J. B. (1994). Anticipated ongoing interaction versus channel effects on relational communication in computer-mediated interaction. Human Communication Research, 40, 473-501.

Walther, J. B. (1996). Computer-mediated communication: Impersonal, Interpersonal and Hyperpersonal Interaction. Communication Research, 23, 3-43.

Walther, J. B. (1998). Leaving virtuality: Theoretical dynamics in the movement from online to offline relationships. Paper presented at the Life in Cyberspace conference, Waltham, MA.

Walther, J. B., & Burgoon, J. K. (1992). Relational communication in computer-mediated interaction. Human Communication Research, 19, 50-88.

When spouses go astray online. (1999, July 21). The Los Angeles Times, p. 3.

Zakon, R. H. (1996). Hobbes’ Internet Timeline. [Online]. Available: http://info.isoc.org/guest/zakon/Internet/History/HIT.html

The present study has continued the empirical exploration of the increasingly popular phenomenon of online dating. In particular, it has looked more closely into the process of development of online relationships and at the types of events that participants in these relationships report as being important. It found that though there are some interesting unique features in online romances, many characteristics of such relationships are relatively similar to those of their face-to-face counterparts. This study also successfully adapted Huston et al.’s (1981) Retrospective Interview Technique (RIT) into an online format. For better or for worse, computer technology is now a mainstay in both our work and personal worlds. It is important that as technology develops and takes on new forms, researchers continue to explore its effects on our lives and also employ it to further their work in beneficial ways.

The purpose of this study was to examine what online romantic partners experience, and how their experiences compare with those of online couples. Our first goal was to develop a typology of turning points for online romantic relationships and to compare these turning point types with those generated for offline romantic relationships. Overall, it was found that the categories originally proposed by Baxter and Bullis (1986) were relevant to both online and offline groups in this study. Two new categories, First Face-to-Face Meeting and Significant Change in Media, also emerged from this sample.

There were three significant differences in turning point frequencies between the online and offline groups. First, First Face-to-Face Meeting was mentioned almost exclusively by the online group. This suggests that as, Hypothesis 2 proposed, this type of event makes up an important turning point category for online relational partners. It can be posited that the reason for this importance is that the first face-to-face (FtF) meeting embodies the symbolic final step in the gradual process of “coming out of the electronic closet” proposed by Walther (1998). This symbolic step seemed to hold important psychological significance to the relational partners. One participant commented, for instance, that simply deciding to meet FtF indicated that “there was some sort of potential.” Another commented that her partner’s act of buying a plane ticket for their first FtF meeting made them be “absolutely sure this was what [they] wanted because of the emotional risk and financial costs involved.”

First FtF meeting may also be important because this is the initial introduction of full shared existence, the ultimate blending between online and offline social worlds. A participant noted, as he anticipated his first FtF meeting with his significant other, that they “had gotten to know each other inside, but how compatible were they on the outside?” Another rejoiced when she discovered that she and her partner were “compatible in person too.” Still another said that his commitment went up upon the discovery that there was “chemistry” in the relationship.

It is important to note that while only one offline participant mentioned her first FtF meeting, we cannot conclude based upon this data that first FtF meeting is not important for the offline group. For the offline participants, first FtF meetings occurred during their first interaction, and in this study, first interactions were not considered turning points as such and participants were not able to describe these events.

The second significant difference in turning point frequencies between online and offline groups was that online participants mentioned Significant Shifts in Media/Exchange of Personal Information twelve times as often as the offline group. This shows that, as Hypothesis 1 states, such events are important to the online romantic partners. In fact, because of the large between-group difference in the frequency with which these events were mentioned, we speculate that shifts in media are more significant to online partners than offline partners. This may be because in online relationships, shifts in media represent significant increases in the number of channels (e.g., vocal, visual) through which a person is able to gain information about his or her partner, whereas for offline partners, there are no such channel increases. We may also speculate based on this information that there is perhaps a sense that for offline relationships, shifts in media (e.g., first phone call) are expected events, whereas for online relationships, it is not necessarily presumed that the relationship will migrate to other forms of communication.

Finally, the online group mentioned the turning point type Exclusivity significantly less than the offline group. One explanation for this difference might be that online relationships are simply less exclusive, perhaps because it is more acceptable to have multiple concurrent romantic relationships online. The finding that online partners did report having significantly more concurrent romantic relationships supports this conjecture. However, this explanation seems to contradict another finding: that online partners often did mention Serious Commitment turning points, some of which assume exclusivity (e.g. engagement and marriage). Another slightly different explanation for the infrequent mention of the Exclusivity category amongst online romantic participants might be that people tend to be unwilling to make exclusivity decisions without meeting FtF. However, this does not match the finding that a majority (59.2%) of the online partners did eventually meet FtF, nor does this concur with the fact that online partners also reported making serious commitment decisions, which might be considered as more “serious” and life-changing than exclusivity. A third explanation might be that the online group was significantly older than the offline sample. Perhaps the notion of exclusivity in dating is less important to the older age group. However, in Bullis, et al.’s (1993) study of turning points in an over-25 sample, no such finding was reported. The reasons for this difference in prevalence of exclusivity turning points remains unclear, but we can conclude that for whatever reason, though online relationships were on average either reported to be more committed or just as committed as offline relationships throughout the 36 months explored in this study, exclusive dating does not seem to be as important a concept to online romantic partners as to offline romantic partners in general.

Hypothesis 1 proposed that Significant Change in Media/Exchange of Personal Info would be an important turning point category and would be related with a large commitment change for online partners. Hypothesis 2 further stated that First Face-to-Face Meeting would be the most important and lead to the largest change in commitment. These hypotheses were not supported. While both turning point categories appeared to be relatively important to online participants, (Significant Change in Media was the sixth most frequently mentioned turning point type, and First Face-to-Face was the third), Significant Change in Media was not as highly ranked as anticipated, and First Face-to-Face was not ranked first. However, the differences between the top three frequencies were extremely small (total of only 6 mentions separated First Face-to-face from Get to Know Time, that was ranked first). In addition, for the two most commonly discussed categories, Get to Know Time and Passion, multiple occurrence within the same relationship is possible, while on the other hand, a person can only meet another for the first time once.

In addition, First Face-to-Face and Significant Change in Media were not related with large commitment changes as anticipated. First Face-to-Face turning points did not lead to the largest commitment changes, as was hypothesized, but were only ranked sixth, and Significant Change in Media turning points were ranked ninth. In addition, the average levels of commitment change related to these categories were not significantly different from those of any other turning point types. This general similarity in relation is comparable to the results reported in Baxter and Bullis (1986), who found that though positive events were related to significantly different commitment changes than negative events, there were no significant differences within the positive and negative groups. These unanticipated small commitment changes may be due to the fact that by the time decisions to change media or to meet face-to-face are made, commitment levels in online relationships are perhaps already so high that they cannot go much higher. Instead, such turning points may just confirm existing high levels of commitment levels. This proposal seems to match the finding that though Serious Commitment events such as marriage and moving in together are commonly seen as necessitating very high levels of commitment, they actually were not related with large commitment level changes (this category was ranked only thirteenth for relation with commitment change). Perhaps for such events too, commitment levels are already so high that by the time the events occur, commitment cannot increase much further.

In such cases where commitment levels are already very high, there is a long way to fall. Cooper and Sportolari (1997), for example, proposed that first FtF meetings would trigger the “bust” in the “boom and bust” pattern to which they suggest, online romances are prone. In this sample, however, first FtF meetings generally were related with increases in commitment. Then again, it is noted that only intact romantic relationships were studied here. It is possible that those online relational partners who did experience the “boom and bust” effect experienced it so strongly that their relationships completely dissolved, leaving the partners ineligible to participate in this study.

Finally, the relatively small commitment changes triggered by the First Face-to-Face Meeting and Significant Shifts in Media categories seem to indicate that these turning points are just like any other and are not experienced as particularly intense. It is posited that just because these two categories represent unique turning points for online romantic partners does not mean that they are important to the exclusion of other turning points. Other more “traditional” events did not cease to be important to online relational partners just as they are to FtF partners. This appears to emphasize Parks and Floyd’s (1996) and Parks and Roberts’ (1998) findings that though online relationships deal with some unique circumstances, they are otherwise relatively normal in character.

The final hypothesis was formulated upon Walther’s (1996) premise that CMC can produce a “hyperpersonal effect.” Specifically, Hypothesis 3 stated that the change in reported level of commitment in online romantic relationships will be faster than that of offline romantic relationships at the beginning of relational development, but that this rate of commitment increase in online relationships will eventually taper off and match the rate in offline romances. Figure 4 seems to partially confirm this hypothesis. It shows that the commitment levels in online relationships tend to begin higher than those in offline relationships and continues to be higher until around 20 months into the relationship. At that point, reported commitment levels in offline relationships catch up with online levels. In addition, when we tested to see whether the mean commitment levels between the two groups were significantly different at each month, we found that all of the significant differences were within the first twelve months of the relationship. Between months 12 and 18, there were a few differences that approached significance. However, after that, none of the differences even approached being significant. This statistically confirms our visual observations that after around the twentieth month into the romantic relationship, commitment rates reported by offline participants begin to match those in the online group. However,

Hypothesis 3 also states that, not only will commitment levels be higher during the beginning of the relationship for online partners, the rate of increase in commitment would also be higher. When we observe Figure 4, there is no such visible higher rate of increase: Commitment levels reported by the online group begin higher and continue to be higher, rising at a parallel rate as those in the offline group until the rate of increase tapers off at around month 15, gradually allowing the offline group’s commitment levels to catch up. This suggests that though there are perhaps aspects of hyperpersonal-ness demonstrated here, but the size of the effect is not as substantial as was expected. This may have been because much of the rationale behind Walther’s (1996) hyperpersonal theory is based upon the concept that computer mediated communication (CMC) is both asynchronous and sparse in the number of cues that it offers. However, most online partners in this study predominantly communicated using forms of CMC that are synchronous. For example, 38.8% of the online participants reported that during their first two months of interaction, their predominant form of communication was instant messaging, and 23% reported that they mostly used Internet chat. In fact, several participants mentioned frustration over the inability to coordinate their schedules with their partners in order to communicate online or noted their disappointment when their partners were not logged on when they expected them to be. These are the sorts of entrainment issues that Walther (1996) suggested are absent in online relationships and that this absence would contribute toward the hyperpersonal effect. In addition, though most forms of CMC that participants used continued to be relatively cue-sparse (i.e., no vocal or visual cues), several participants did mention setting up microphones and webcams. These new features of technology that make CMC both synchronous and richer in cues may explain why the hyperpersonal effect, though seemingly present here to an extent, was not as strong as predicted.

There are several limitations to this study. First of all, there were some significant demographic differences between the online and offline groups (e.g., in age, sex, and marital status). This makes it difficult to say conclusively that relational differences between the two groups are due to differences in communicative medium rather than demographics. In addition, some similar demographic differences were found between the two subsets of the online group: those who had previously been in a FtF relationship and those who had not. Future research should seek to make the same online-offline comparisons using populations that are demographically more similar, and to explore how previous romantic experiences affect current relational characteristics.

Another limitation is that Walther’s (1996) hyperpersonal effect, which he defined as when CMC is experienced as more intimate and interpersonal than similar FtF communication, is conceptualized in the present study simply as a higher rate of commitment increase. This is perhaps an over simplistic reduction of this concept, and future studies should look more directly at the interpersonal quality of the relationships throughout their development. This could be done by doing a close study of the communication that participants report having during each turning point.

The fact that only intact romantic relationships are studied here is also a weakness. By looking at unsuccessful relationships as well, we would be able to more fully explore the effects of the first FtF meeting in online relationships. It might be interesting, for instance, to explore what the differences are between first FtF meetings that lead to relational success and those that lead to relationship dissolution. Additionally, including participants in previous relationships would also allow us to explore general development differences between successful and unsuccessful online relationships. For example, it may be interesting to compare commitment increase rates between successful and unsuccessful groups.

In this study, we found that around the 20 th month of relational development, commitment levels reported by offline romantic partners began to match those of online partners. Future studies should explore whether this period of time around the 20 th month is critical, and if so, what its significance is. This could be done by looking more closely at the types of turning points that typically occur during this stage of relational development.

Hypothesis 3 posited that the change in reported level of commitment in online romantic relationships will be faster than that in offline relationships at the beginning of relational development but that this rate of commitment change in online relationships would taper off, allowing the rate in offline romances to catch up. To test this hypothesis, curves representing an average relationship trajectory were drawn for the online and offline groups. To calculate the average graph for each group, the commitment scores for each respondent at the beginning of each month were calculated. This was done by extrapolating the level of commitment at each month using the slope of the line connecting the immediately preceding and subsequent turning points. For example, if a the first turning point occurred at month 1 and at a commitment level of 10%, and the second turning point occurred at month 3 with a commitment level of 30%, the investigator assumed linearity and extrapolated that at month 2, the level of commitment was 20%. These monthly commitment scores were then averaged for each group and plotted on a graph to form the average curves. To maximize accuracy, any data from participants who generated less than two turning points were excluded. The two average curves generated are displayed in Figure 4.

To figure out where significant mean differences between online and offline groups in levels of commitment lay, t-tests were done for commitment levels at each month. Results are shown in Table 5.

gure 4: Mean Reported Commitment Levels Over Time for Online and Offline Groups

Table 5: Differences in Mean Monthly Commitment Levels between Online and Offline Groups

Hypotheses 1 proposed that turning points involving significant shifts in media use and exchange of real-life information would be important to online partners and be related to large changes in reported commitment levels. Hypothesis 2 further posited that first face-to-face (FtF) meeting with one’s significant other would be the most important turning point to online partners and would lead to the largest changes in reported commitment levels.

For each turning point, participants had been asked to plot the level of commitment immediately following the occurrence of each turning point. Therefore, the commitment change related to each turning point was calculated by first taking the commitment level of each turning point and subtracting the commitment level of the previous turning point. These difference scores were then averaged for each turning point category. Categories with an average commitment difference score greater than zero were viewed as positive turning points, whereas those with an average difference score of less than zero were viewed as negative. Table 3 illustrates average commitment change levels by turning point type for all participants, from most positive to the most negative. Results were similar to those found by Baxter and Bullis (1986): For both online and offline groups, all categories were positive, with the exceptions of External Competition, Relational Problems, and Negative Psychic Change.

Table 3: Means and Standard Deviations for Change in Reported Relationship

To determine the absolute levels of change in commitment each turning point type was related to, the absolute values of the commitment change scores for each turning point category were averaged. Mean absolute values and rankings for each category are shown in Table 4. A between-group analysis of variance of overall differences was conducted on the various categories, with absolute change in commitment level as the dependent variable. The main effect on turning point type was found to be significant (F = 3.15, df = 15, p < .001).

Table 4: Means, Standard Deviations, and Ranks for Absolute Change in Reported Commitment Level by Turning Point Type for Online Group

However, Scheffé’s post-hoc tests demonstrated that the only significant individual difference in absolute difference in commitment was between the Passion (M = 60) and Quality Time categories (M = 10.13) (F = 23.58, p < .05). Some seemingly large differences, for example between Exclusivity (M = 39.38) and Quality Time (M = 10.13), didn’t reach significance, probably because of smaller sample sizes.

The first two research questions focused on developing and comparing typologies of turning points for the online and offline samples. Four hundred and fifty-four turning points were identified by the online group and 452 by the offline group, resulting in a total of 906 turning points. Two new categories were added to the 14 supra-categories derived inductively by Baxter and Bullis (1986) to account for additional reported types of events, to make a total of 16 turning point types. A comparison of frequency and percentage distributions between the online and offline groups is reported in Table 1. A rank-difference correlation (Spearman-Rho) was done to compare the rankings for both groups and it was found that the they were significantly correlated (r = .52, p <.05). An additional test of correlation was calculated to compare the turning points of online participants who had previous offline romantic experience with those who had not. The turning point frequency rankings for the two sub -groups were found to be highly correlated (r = .80, p <.001). Because of the strong correlation between the turning point frequencies of the online group with previous FtF experience and the group without, it was determined unnecessary to explore any further differences between these sub -groups. Table 2 shows the turning point frequencies and percentage distributions for both online sub-groups.

Table 1: Distribution of Turning Point Types in Online and Offline Groups

The supra-type Get-To-Know Time was mentioned most frequently by respondents in both online and offline groups (see Table 1). This finding replicates those in Baxter and Bullis’ (1986) and Bullis et al.’s (1993) studies. This category encompassed occurrences that facilitate getting to know one another better, such as first dates, and spending time together enacting various activities (e.g., sports, recreation, studying). To account for the data in this study, an additional sub-group of turning points, labeled “intense conversations,” was also attributed to this category because it was found that many couples spent time getting to know each other by simply talking with each other (either FtF or on the telephone), rather than doing particu lar activities. Intense conversations included events such as “we stayed up all night and talked about our childhoods” or “we discussed the things that are important to us.”

Table 2: Distribution of Turning Point Types in Online Group with no Previous FtF Relationships and Online Group with Previous FtF Relationship

The second most common turning point type mentioned by both groups was Passion. This referred to events that involved any sort of physical or emotional expression of affection between partners. Included within this category were first sex, first kiss, first admissions of love or “feelings for one another,” and love-at-first-sight experiences. In addition to these events described by Baxter and Bullis (1986), events centered around “flirting” were also considered here as belonging in this category.

The supra-type Quality Time comprised of special events for the couple to appreciate one another and their relationship. Most frequently, this category included getting away from the day to day (vacations together), meeting the parents and/or other family members, seeing where the other grew up, relationship anniversaries, and special holidays (e.g. Valentine’s Day, Christmas, New Year’s).

Serious Commitment was the next most commonly mentioned category for both groups after Quality Time. Such serious commitment events included making marital plans (talking about getting married, getting engaged) and moving in together. It was found that a significant number of participants mentioned moving to live closer to one another. While this was different from moving into the same household together, these events were still included in the Serious Commitment category because this sort of life-change was considered to require significant commitment to the relationship, especially as many participants mentioned moving from a significant distance away to be in the same city or country as the relational partner. For example, one participant wrote about a move from Alabama to Arizona, and another talked about her imminent move from New Zealand to be with her significant other in England.

The category Relational Problems is very similar to that named “Disengagement” in Baxter and Bullis (1986) and Bullis et al. (1993). It encompassed fights, conflicts, disagreements, and other relational problems such as let-downs and betrayal (though exclusive of external competition). The title for this category was changed because the word “disengagement” and the description of “any de-escalation in relationship” did not seem to adequately describe many of the relational problems that were mentioned. For example, many participants wrote about fights that were perhaps bitter and heated but yet resulted in a rise rather than de-escalation in commitment. Some explained that though they were emotionally hurt by such conflicts, they believed that the conflicts demonstrated the couple’s determination to work through problems rather than to ignore them, thus their commitment to the relationship rose after the event. This category’s title was also changed to better accommodate mentions about stressful events that were painful, but resulted in an increase in commitment, such as having a miscarriage or making the decision to have an abortion.

Of the relationships that went through relational problems that led to a complete dissolution or break-up, a proportion was restored. The category Making Up encompassed such occurrences when partners got back together.

Exclusivity was the category of turning points encompassing making a joint decision to engage solely in the current romantic relationship and the breaking of romantic involvement with all others except the current partner. Also included in this group were events surrounding the process of becoming an “official” couple (either a decision made by the partners or being recognized by others as being officially a couple). The frequency of turning points in this category was significantly different for online and offline groups, with the offline partners (f = 33) mentioning this category more often than the online partners (f = 13) (c 2 = 8.70, df = 1, p < .01).

The category Physical Separation/Losing Contact involved separations due to vacations, school breaks, and overseas trips, rather than separations due to break-up of relationship. Also added into this category was loss of contact, usually due to inability to access communication lines (e.g. no Internet access). Related to this category was Reunion/Reestablishment of Contact. This concerned events when partners were reunited after physical separation or when communication was reestablished.

External Competition referred to the arrival of a new rival (a third person) who is reported as competing for the affection of one of the partners, the return of an ex-significant other, or the interference of competing non-romantic demands such as work, school, friends, or children.

Ranked tenth for both groups was the category Sacrifice. This involved crisis help or support—assistance by either partners when the other was experiencing a personal problem—as well as the giving of favors or gifts by either partner.

Categories Positive Psychic Change and Negative Psychic Change both refer to intra-psychic fluctuations in a person’s attitude or feeling toward the relationship that were not reportedly triggered by any external event.

Two new categories were added to take into account events that were often mentioned, particularly by online romantic partners. Significant Shifts in Media/Exchange of Personal Information referred to the expansion of communication to new types of media (e.g., first phone call, started exchanging emails, exchanged photos), as well as the exchange of personal information (e.g., exchanging phone numbers or email addresses, revelation of real age). This category was mentioned significantly moreoften by the online (f = 37) group than the offline group (f = 3) (c 2 = 28.90, df = 1, p < .001). Also added was the category First Face-to-Face Meeting, which included both planning for the first meeting, as well as meeting for the first time. This category was mentioned significantly more often by the online (f = 59) than the offline group (f = 1) as well (c 2 = 56.07, df = 1, p < .0001).

Because there was some concern that online romantic partners who had never been in any prior offline romantic relationships might be considerably different from those who had previously had offline relationships, several tests were done to explore possible differences between these sub-groups. It was found that 72% of online participants had previously had a face-to-face (FtF) romantic relationship. The two sub-groups included similar proportions of men and women, were slightly different in racial make-up, with the group having no previous FtF romantic experience including a larger proportion of Asians (28.6% vs. 5.5%), and slightly fewer Caucasians (60.7% vs. 79.5%). The rest of the racial differences were hard to interpret because of very small group sizes and because some categories actually had no representatives at all. In addition, the typical marital status of participants in the two sub-groups was different (c 2 = 9.30, df = 2, p = .01), with the largest difference being the online-only group’s larger proportion of divorced individuals (33.3%). The group with past FtF romantic experience was also significantly older (M = 28.90, SD = 9.02) than the group with no past FtF experience (M = 24.86, SD = 7.71) (t = 2.10, df = 99, p < .05).

Several general comparisons of the online and offline groups were made. Both groups consisted predominantly of women, but sex differences were more common in the online group. The offline group had twice as many males (30.4%) as did the online group (17%) (c 2 = 5.17, df = 1, p <.05). The online group was also older (M = 27.72, SD = 8.81) than the offline group (M = 22.96, SD = 6.63), and this difference was statistically significant (t = 4.43, df = 202, p < .001). The two groups’ racial make-up was not significantly different, however, their marital status was (c 2 = 21.25, df = 1, p < .001). This difference was primarily due to the much greater number of divorced participants (n = 24) in the online group than in the offline (n = 3). The online group (M = 21.09, SD = 15.34) also reportedly spent significantly more hours on the Internet than did the offline group (M = 13.98, SD = 18.66) (t = 2.99, df = 204, p < .01). In addition, online partners reported having significantly more concurrent romantic relationships (M = 1.42, SD = 1.26) than offline partners (M = 1.14, SD = .61) (t = 2.02, df = 203, p < .05).

For ease of grouping, each reported turning point was printed out onto separate slips of paper. Using Baxter and Bullis’ (1986) 14 main categories of turning points as a beginning framework, the researcher coded 10% of all the turning points from a mixture of both online and offline surveys. When turning points did not fit into the existing categories, changes were made to the Baxter and Bullis (1996) codes or additional categories were added (see Table 1). The researcher and another coder then independently coded 20% of the turning points, and reliability of the coding was assessed. The level of intercoder reached was similar to those in Baxter and Bullis (1996) and deemed adequate (Cohen, 1960, kappa = .76), therefore the primary investigator coded the remaining turning points.

Data collection took place exclusively on the Internet. Potential participants were directed to an online survey located on a web site. The survey began with an explanation of the purpose of the study, qualifications to participate in the study, and an assurance of confidentiality.

In Part 1 of the survey, participants were asked for basic demographic information, information about Internet usage, general information about the dating relationship on which they would be reporting, as well as information about past romantic involvement. Figure 2 is a screen-capture of what the participants saw in this section of the study.

Part 2 of the survey consisted of a modified version of Huston et al.’s (1981) Retrospective Interview Technique (RIT) that was used in Baxter and Bullis (1986) and Bullis et al. (1993). The RIT asks respondents to identify all of the turning points in their relationship since time of first interaction and to plot these points chronologically on a graph. After completing Part 1, participants came to a web page showing a graph (Figure 3). The x-axis of the graph was scaled automatically for the length of the relationship reported in Part 1, and it was divided up into monthly intervals, beginning at the date the relationship began, and ending with the current date. The y-axis measured the interviewee’s sense of commitment in the relationship ranging from 0% to 100%. As in Baxter and Bullis (1986) and Bullis et al. (1993), “commitment” was undefined and left to the participants’ interpretations.

Figure 2: Screen Shot of Part 1 of Survey

Participants were then asked to use the computer’s mouse to plot five major relational turning points on this graph. To plot a turning point, the participants used their cursor to locate the spot on the graph representing the time the event occurred and their perceived level of commitment immediately following the event. To record the turning point, participants simply clicked on the mouse. A window then appeared asking the participant to label that event. Adjustments could be made to the turning points until participants clicked on a button indicating that they have finished identifying the relationship’s turning points. Once this was done, a line was automatically drawn linking all five turning points, the first interaction, and the current date.

Figure 3: Screen Shot of Turning Point Graph

In the next part of the study, participants were asked for additional information about each of the identified turning points. The graph with the turning points remained on the screen so that participants could refer to it. For each turning point, the participant was asked to answer the following questions (modified from Baxter & Bullis, 1986; and Bullis et al., 1993): (a) Please explain in detail what happened during this event; and (b) Why did this event result in change in the relationship’s level of commitment?

Once this section was complete, a few more questions regarding the relationship were asked and the study was complete. Participants were thanked and informed that if they would like to receive the results of the study or be entered into the prize drawing for the gift certificate, they could submit their email address. This address was kept in a separate file from the responses to ensure anonymity. The entire survey typically took about 20 minutes to complete.