Towards a​​ multidimensional methodology for quantification of romantic​​ strife:​​ Developing​​ objective​​ tools for evaluating​​ the​​ utility of​​ relationship maintenance

K. Terwilliger-Hardy, D.M. Young, & A.J. Hardy

Abstract

The pursuit and maintenance of romantic pair bonds can be a time-consuming and frustrating endeavour for many humans.​​ All​​ long-term pair bonds​​ are eventually​​ met with​​ challenges that must be overcome at the expense of harmony in order to maintain the bond.​​ Observational​​ data across history support the possibility that a certain degree of romantic strife is necessary and even beneficial to romantic partners. The present study aimed to explore factors that contribute to romantic strife in the interest of determining whether there are domains in which strife may serve a functional role in romantic pair bonds, and the magnitude of strife at which pair bonds become dysfunctional.​​ The implications of​​ findings​​ from three experiments​​ for the development of objective romantic​​ quantification metrics​​ are discussed.

Introduction

Since the straightening of the first​​ Homo erectus​​ spine, much of day-to-day human existence has centered​​ on​​ the pursuit and maintenance of romantic pair bonds. Indeed,​​ many​​ art forms in contemporary culture are focused, directly or indirectly, on the relative consideration of pragmatic costs versus​​ benefits​​ in romantic partnership (cf.​​ influences in literature and music such as​​ Romeo & Juliet,​​ Pride and Prejudice,​​ “What’s Love Got To Do With It,”​​ “Love Will Keep Us Together,” & “99 Problems,” to name but​​ a few). It would seem, from the iconic status attained by such media, that​​ the ability to overcome​​ a certain degree of strife is​​ indicative of the​​ perceived​​ profundity and depth​​ of​​ a union between​​ romantic partners. On the other hand, strife in human pair bonds may also lead to decreased life satisfaction (Terwilliger-Keyes & Keyes, 1985; 1991; 1994; see also Terwilliger-Schultz, 2009​​ for review), as well as increased levels of blood cortisol (Terwilliger-Keyes​​ & Schultz, 2005), which is associated with​​ a wealth of negative health​​ outcomes.

The present study aims to assess the relationship between strife and romantic love in human pair bonds. If such a relationship exists, it​​ is the authors’​​ contention​​ that​​ most individuals​​ would be well served by a tool​​ that​​ helps reduce​​ the number of hours spent in preoccupation with romantic problems​​ (see Young, Keyes & Schultz, 2014 for review). As such, one goal of the present study is to develop​​ a novel methodology for quantifying​​ the​​ optimal​​ strife/love balance​​ in human relationships.​​ By developing this tool we hope to​​ provide​​ research-based, concrete metrics that allow individuals to know when it’s time to​​ move on. This will save many adults the costs—financial, physical, and psychological—of unnecessarily enduring​​ unhealthy​​ levels of romantic strife, thereby facilitating the functioning of a productive society unpreoccupied with unsolvable problems (see Schultz & Young, 2011 for review).

To this end, we recruited five romantic couples to participate in a five-year longitudinal study​​ involving several experiments.​​ To​​ simplify the potential interpretations of our results,​​ we selected​​ only​​ monogamous​​ couples with no children, who had not been together longer than five years at the onset of the study.​​ We present results from three experiments​​ and discuss the implications for the development of our multidimensional love quantification tool.

Experiment 1

 Experiment 1 was designed to assess​​ the relationship between “partner similarity”​​ and romantic strife.​​ We predicted that romantic partners who are highly dissimilar would experience more strife. Further, we expected that partners who were too similar would also experience relatively more strife,​​ whereas partners of medium similarity would experience less strife​​ (personal communication, Terwilliger-Hardy, 2015).​​ 

Methods

Participants

 Five romantic couples were recruited via community advertisement to participate in the study.​​ All couples were without children and had been together for fewer than five years (with a range of 2.1 to 4.8 years).​​ Couples​​ #2 and​​ #5 terminated their romantic partnerships prior to the end of the five-year study period, but continued participation in the study, with modifications where necessary.

Stimuli​​ & Procedure

 The Terwilliger-Keyes Romantic Partnership Battery (RPB; Terwilliger-Keyes & Keyes, 1996) was administered. The RPB is a tool developed to assess satisfaction in romantic partnerships, and includes such measures as the​​ Romantic Activity Assessment Scale (RAAS; Terwilliger-Young & Young, 1980),​​ the Big 5 Personality Inventory,​​ Satisfaction With Life (SWL), and sex-specific Golombok Rust Inventory of Sexual Satisfaction (GRISS), among others. For a full listing of the battery questionnaires and​​ validation for use in assessing romantic satisfaction, see Terwilliger-Keyes and Keyes (1996).​​ 

Results​​ & Discussion

 For each questionnaire within the RPB, we computed a correlation of scores between romantic partners. We​​ then​​ computed the average of correlations for each couple in order to ascertain an aggregate measure of overall partner similarity,​​ hereafter referred​​ to as the Pea-Pod Index (PPI). The PPI estimate​​ ranges from 0 to 1,​​ where higher scores indicate more similarity between partners.​​ 

We used the PPI as a predictor for scores on the Terwilliger-Schultz Romantic​​ Exasperation​​ Scale (RES; Terwilliger-Schultz & Schultz, 2004), a​​ measure designed to assess romantic strife. A​​ lower​​ RES score indicates, to put it plainly, a higher degree of “lovey doveyness” (i.e., the feeling that one’s partner can do no wrong),​​ while a​​ middling score indicates a neutral level of​​ strife, and a​​ higher​​ score​​ indicates more​​ exasperation (i.e.,​​ romantic strife).​​ 

Figure 1.​​ Plot of PPI and RES scores for each couple.

We predicted a U-shaped relationship between the PPI and strife. However, contrary to our expectations the data appear to conform to an inverted-U relationship​​ (see​​ Figure 1), in that partners who are either highly similar or highly dissimilar​​ report less strife, while partners of middling similarity report​​ more strife. However, these results​​ should be interpreted with caution, as​​ PPI scores​​ did not reliably predict relationship​​ satisfaction.​​ 

Our longitudinal analysis​​ also​​ revealed no consistent pattern​​ between PPI and strife.​​ Exasperation​​ scores tended to increase for couples over time, as supported by previous work​​ (Terwilliger-Schultz & Hardy, 2013). PPI scores also fluctuated, with some couples increasing​​ in​​ similarity over time, and others decreasing. Of the two couples that terminated relationships, one demonstrated an increase in PPI, but a decrease in​​ exasperation​​ between years one and three of the study, while the other demonstrated a decrease in PPI​​ but increase in​​ exasperation​​ between years one and four of the study.​​ 

Exasperation​​ also did not predict relationship​​ satisfaction, nor did magnitude in change of​​ exasperation​​ over time. This is possibly due to the small sample size.​​ With enough couples, it is possible that the rate of increase in exasperation over time would be a significant predictor of relationship​​ satisfaction.​​ Further research is needed to determine the​​ utility of assessing​​ partner similarity​​ as pertaining to​​ strife in romantic pair bonding.

Experiment 2

In this experiment we aimed to determine whether romantic partners’​​ knowledge​​ about each other​​ influences the amount of strife in relationships, and whether such an influence may have implications for relationship success. Previous work by Terwilliger (1978) suggests that romantic partners report more relationship satisfaction when they are with someone who​​ demonstrates​​ memory​​ for​​ the​​ inconsequential details​​ of their lives.​​ It is not unreasonable to expect that feelings of​​ “being interesting”​​ to one’s romantic partner facilitate positive attachment styles, thus bolstering romantic love. In this particular case, we would expect that romantic strife should be low when partners​​ commit to remembering inconsequential details about each other, and that​​ strife in this context may be functionally exploited​​ to increase relationship satisfaction. In other words, one partner’s strife may​​ induce a greater amount of effort​​ by the other partner to maintain the bond.

Methods

Stimuli & Procedure

 This experiment was modeled after​​ The Newlywed Game. Testing was performed once per year for the duration of the study, and all couples answered the same set of​​ ten​​ questions in a given year.​​ 

 Partners were seated in separate rooms, and each provided answers to the set of questions. In this pre-testing period, each individual was also asked to rate their confidence​​ out of ten​​ regarding the likelihood that their partner would correctly identify the answer they had given to each question.​​ Finally, couples were reunited for the quiz session, where they alternated attempts to guess at what their partner’s answers would be. All sessions were videotaped.

Results & Discussion

 In order to assess the influence of partner-knowledge on​​ romantic​​ strife, we conducted a multiple regression using both performance accuracy and confidence ratings to predict couples’ exasperation scores. The​​ results showed​​ a significant effect of​​ accuracy:​​ couples who performed better on the quiz session experienced​​ less strife​​ on average.​​ However, inspection of individual​​ video​​ sessions demonstrated that this effect was​​ only observed​​ when​​ both​​ partners​​ performed well. On​​ several​​ occasions, the researchers​​ were forced to intervene​​ in order​​ to terminate elaborate screaming matches (n = 7) and​​ vase-smashing​​ (n = 1)​​ when one partner significantly outperformed another.​​ Further,​​ there were​​ numerous​​ instances​​ of crying​​ during the quiz period, indicating that​​ taking an​​ average​​ accuracy​​ between partners​​ may not accurately capture overall strife.​​ 

 Additionally, this finding was​​ qualified by an interaction between accuracy and confidence​​ ratings.​​ Strife​​ was​​ higher in couples​​ that​​ had​​ low confidence​​ in their knowledge of one another,​​ but performed with high accuracy.​​ On the other hand,​​ strife was​​ lower in couples​​ that​​ had high confidence,​​ even when accuracy was low.​​ Contrary to our prediction, these results suggest that strife is influenced mainly by the discrepancy between what a person expects their partner to know and what their partner actually knows.​​ In particular, confidence in one’s partner’s knowledge may be more important​​ for regulating relationship strife​​ than actual knowledge.​​ 

Participants’ performance accuracy did not change significantly over time​​ (see Figure 2). This was contrary to our expectation, and suggests that individuals do not necessarily invest effort in continuing to encode inconsequential details about romantic​​ partners, but rather plateau in their knowledge at some point relatively early (i.e., < 2.1 years) in a new relationship. On the other hand, confidence ratings significantly increased with time, such that the longer partners were together, the more detailed information each partner​​ expected​​ the other​​ to​​ know​​ (Figure 2). Correspondingly, strife increased significantly over time as a function of the growing discrepancy between partners’ confidence and actual performance.​​ It would appear that as relationships progress, partners may become increasingly sensitive to perceived lapses of memory, regardless of confidence in one another’s knowledge.

Figure 2.​​ Longitudinal comparison of quiz performance and confidence levels. Larger discrepancies between confidence and accuracy were associated with higher magnitudes of strife​​ (data not shown).

 In order to assess the influence of strife on relationship success, we performed an additional logistic regression to predict relationship outcome (Success vs. Fail) as a function of​​ exasperation​​ scores and​​ confidence – accuracy difference scores.​​ This analysis revealed​​ that relationships were likely to fail when accuracy exceeded​​ confidence and strife was low, and did not fail when confidence exceeded accuracy and strife was high. These​​ groundbreaking​​ data suggest that strife may be a positive contributor to relationships only when partners’ confidence in one another’s interest is high, an entirely​​ novel insight which may prove crucial​​ not only​​ to the development of a love quantification metric, but to expanding the broader field of relationship research in general.

Experiment 3

This experiment was designed to assess the​​ relationship between​​ physiological arousal tendencies (i.e., lust),​​ romantic​​ strife,​​ and relationship satisfaction.​​ It is widely known that in monogamous pair bonds, there is great potential for strife to be introduced when partners express inappropriately-directed lust​​ (lust expressed​​ at a​​ non-optimal​​ time or towards a non-partner individual).​​ We predicted​​ that​​ monogamous​​ partners scoring high in relationship satisfaction would experience the least arousal to​​ pornographic​​ images (i.e., non-partner images), while partners with low relationship satisfaction would experience higher arousal. Further, we predicted an interaction between arousal and relationship satisfaction such that inappropriately-directed​​ lust​​ would cause more strife in “happy” couples and matter less to couples who are already unhappy​​ (but see Schultz & Young, 2004 for counter-arguments).

Methods

Stimuli & Procedure

 All participants were tested once per year throughout the duration of the study.​​ In order to measure physiological arousal, participants were fitted with penile plethysmograph or vaginal photoplethysmograph while viewing a series of pictures. The plethysmograph measures changes in blood flow to the genitalia, thus providing the most direct possible measure of erotic arousal.​​ This approach, while somewhat invasive, is preferred over traditional self-report metrics of arousal as it circumvents the possibility that individuals will​​ underreport​​ arousal levels for fear of embarrassment or​​ retaliation.

 Pictures included​​ ten images from each of the following categories: heterosexual pornography;​​ homosexual pornography​​ (male and female); masturbatory scenes​​ (male and female); and neutral images (e.g., clocks, coffee mugs) to provide a physiological baseline against which to compare arousal. Participants were​​ tested individually to​​ prevent interference​​ effects​​ from​​ the presence of​​ romantic partners. In each session, pictures were​​ presented​​ one at a time in random order for ten seconds​​ each, with a thirty second break in between each picture. Participants were told that their memory for the images would be tested to facilitate attention to the images.

Results & Discussion

 Arousal levels elicited by​​ all pornographic​​ image​​ categories​​ (compared to neutral baseline) were mixed, with some participants​​ demonstrating​​ high arousal and others​​ demonstrating​​ virtually no​​ arousal​​ at all.​​ Counter​​ to expectations, arousal was​​ not​​ related to relationship satisfaction or strife​​ on average.​​ Arousal levels also fluctuated unpredictably from session to session during the five-year study period.

 Intuitively it would seem perplexing that tendency toward sexual arousal bears no association​​ whatsoever​​ with romantic partnership variables. In order to further clarify the nature of our findings, we examined data separately for each couple.​​ We describe here our observational findings for couples #3 and #5, with the lowest and highest scores on relationship satisfaction, respectively.

Arousal findings for couple #3

 One partner,​​ AP,​​ demonstrated high levels of arousal to​​ all types​​ pornographic images relative to neutral baseline, while AS, the​​ other​​ partner, demonstrated near-zero​​ or even negative​​ arousal​​ across all pornographic categories. However, when we examined his data more closely, we learned that this finding was an artifact arising from AS having experienced unusually high levels of arousal​​ in the neutral image category. Both partners scored high on sexual satisfaction, and while relationship satisfaction was lowest​​ compared to the other couples in the study, the level of romantic strife was moderate and stable over the five-year period. By the end of the study period, the partners in couple #3 had married.

Arousal findings for couple #5

 In contrast with couple #3, couple #5 scored the highest on measures of relationship satisfaction throughout the four-year period of the study prior to termination of the partnership. Strife​​ levels​​ were volatile, fluctuating between the lovey-dovey and exasperated extremes of the scale.​​ Partner PS rated sexual satisfaction as high, while partner ED rated sexual satisfaction as low-moderate.​​ Both partners demonstrated a​​ moderate arousal level to all pornographic image categories, with a slight increase for the homosexual category​​ (females > males).​​ Interestingly, both​​ partners​​ demonstrated a high degree of similarity in arousal response across categories, suggesting somewhat unexpectedly that a tendency to become aroused in the same situations does not predict relationship success.​​ Indeed, the results from couple #3 would suggest that most beneficial to long-term pair bonds is a high level of non-specific​​ arousal, allowing us to conclude that there may, after all, be no such thing as inappropriately-directed lust​​ in the context of a romantic pair bond​​ (but see Young, Hardy & Schultz, 2008 for additional perspectives).

General​​ Discussion

 In a groundbreaking longitudinal study of relationship strife, we have shown across three experiments that strife may serve a functional role in the pragmatics of maintaining versus terminating a romantic pair bond. In Experiment 1 we demonstrated that similarity between romantic partners may be associated with romantic strife. Further research is necessary to clarify this possible connection, as similarity was unrelated to relationship satisfaction. Experiment 2 revealed that actual​​ knowledge about one’s partner​​ matters​​ less​​ than​​ confidence​​ in partner-knowledge.​​ Relationships​​ succeeded when confidence was higher than accuracy, and strife was high, suggesting that strife may motivate partners to invest effort in maintaining​​ or increasing​​ one another’s​​ confidence​​ (cf. also Terwilliger-Young, 1983).

In Experiment 3​​ we showed that inappropriately-directed sexual arousal does not significantly predict relationship strife​​ in monogamous partners.​​ Conversely, our data suggest that generalized lust​​ may be​​ associated with successful relationship outcomes.​​ This finding was surprising, to say the least, and may suggest​​ more broadly​​ that​​ infidelity bears no influence on the functionality of romantic pair bonds, and that​​ popular culture’s obsession with romantic betrayal does not truly capture human nature​​ (but see Schultz, Keyes & Young, 2013, for counter-argument).​​ Further research is needed on the impact of infidelity.

 As with any exploratory research program, much work remains to be done in order to isolate the most important factors contributing to “tipping points” of romantic strife.​​ We report here a “first pass” at quantification, having identified variables that​​ contribute to strife, as well as one surprising variable (i.e., inappropriate arousal) that does not.​​ 

 Most promising are our findings regarding confidence in partner-knowledge.​​ Of particular interest is the tendency for confidence in one’s partner to increase linearly from year to year, even when provided with evidence (i.e., yearly performance on the quiz) that actual knowledge is static. This suggests strongly that partner-delusion may be an important factor in romantic pair bonds, and is an exciting avenue for future study​​ including functional thresholds of delusion and the development of delusion-maintenance techniques for romantic partnership counseling.​​ 

Acknowledgements.​​ This research was supported by government funding as well as by​​ an​​ anonymous personal donation to the laboratory of Terwilliger-Hardy.

Conflicts of​​ Interest​​ Statement.​​ The contributing authors report that the research was conducted in the absence of any relationships that could be construed as a potential conflict of interest.