Abstract
Because on-again/off-again (on-off) relationships typically maintain their relationships in some form even after relational dissolution, a communication process that may be particularly relevant to understanding these relationships is relational maintenance. Using cross-sectional data, the current study assessed relational maintenance, uncertainty, and commitment to further understand how on-off relationships differ from relationships that do not have a history of renewals (i.e., noncyclical relationships). The sample of 487 participants completed a survey on either their current dating relationship or a postdissolution relationship. As predicted, on-off partners used less maintenance behaviors than noncyclical partners in current relationships, but, contrary to predictions, on-off partners did not use more maintenance behaviors in their postdissolution relationships than partners without a history of renewing. Path models assessing the interrelationships among maintenance, relationship uncertainty, and commitment also suggest relational maintenance operates somewhat differently in on-off and noncyclical relationships. Implications for on-off as well as dating relationships in general are discussed.
Acknowledgements
A previous version of this paper was presented at the 2008 National Communication Association Convention, San Diego, CA.
Notes
1. Because both the length of relationship and time since breakup variables were skewed, we also conducted the comparisons with the variables transformed using their square root. The analysis yielded the same results and thus the results with the raw numbers are presented as these are easier to interpret.
2. For the EFA for current partners: factor loadings for the five items included in the positivity factor ranged from .69 to .77 (all cross-loadings<.17); factor loadings for the five openness items ranged from .69 to .84 (all cross-loadings<.11); and factor loadings for the four network items ranged from .64 to .88 (all cross-loadings<.19). For the EFA for postdissolution partners: factor loadings for the five items included in the positivity factor ranged from .64 to .87 (all cross-loadings<.18); factor loadings for the five openness items ranged from .72 to .81 (all cross-loadings<.14); and factor loadings for the four network items ranged from .54 to .84 (all cross-loadings<.24).
3. Full tables with all path coefficients and their significance levels for all models can be obtained from the first author.
4. A full table of all models conducted to test for specific significant differences can be obtained from the first author.