Psychologists are interested not only in why people fall in love but why they fall out of love. Gottman and Bradbury and Fincham both argue that communication styles lead to the end of the relationships. Hatfield argues that it is more about a lack of equity in a relationship. However, studying why people fall out of love is not so easy.  There are many limitations to research on relationships ending.

Hatfield argued that a lack of equity in a relationship is what leads to marital dissatisfaction.  In otherwise, one or both partners perceive that they are investing a lot into a relationship but getting little back in return.  Regardless of the actual realities of the costs and benefits of the relationship, this perception leads to dissatisfaction.  Hatfield gave questionnaires to over 2000 couples.  She found that those who felt that their relationship was equitable were more likely to predict that they would be together in five years. Those who felt under or over-benefited in the relationship were least likely to think that they would be together in the future. However, this study was hypothetical and did not actually verify whether these predictions were accurate.  This means that although equity may be used to evaluate the current health of a relationship, it may not be a good indicator of the end of a relationship.

Bradbury and Fincham argue that how couples communicate with each other influences the health of a relationship. They found that unhappy couples blame their partners for bad things that happen and don’t give them credit for positive things. The researchers carried out a prospective study to see what role communication might play in the health of a relationship. They found that when couples used positive attribution patterns (when the partner is successful, it is because of him; when the partner fails, it is because of the situation), this was a good predictor of a happy marriage and negative attribution patterns predicted low marital satisfaction or the end of the relationship.  

Gottman also studies the role of communication in relationships.  His research is based on his work with couples who have come to him for counselling.  He observes the couple having an argument and measures their heart-rate, blood pressure and galvanic skin response.  Through the analysis of these sessions, he argues that there are four communication patterns that lead to the end of a relationship: criticism, defensiveness, contempt and stonewalling.

These three studies show some of the key limitations of research on why relationships end. One key limitation is the problem of bidirectional ambiguity.  All of the studies above are correlational in nature.  It is not possible to establish a cause and effect relationship with these variables.  For example, in the Bradburty and Fincham study, does negative attribution lead to the end of a relationship, or do unhealthy relationships lead to negative attributions?  In Gottman’s research as well, it is not clear whether the communication style is the cause of the unhealthy relationship or an effect.  In addition, focusing on a single factor is reductionist and does not address the complex nature of human relationships.

Both B & F and Gottman have done prospective research. However, much of the research on the end of relationships relies on the memory of the participants. Asking couples in distress why they are unhappy may be influenced by a cognitive bias called “peak end rule.” When asked to explain their relationship their memories will be influenced by the most recent events and any really bad days.  For example, they may argue that their partner “never confides in them”, but actually, this is a memory distortion.  As the relationship has deteriorated, the level of disclosure has decreased.   

Although Gottman uses physiological tests to support his research, one still has to question the operationalization of variables.  How does one measure one’s level of contempt?  Or one’s level of disclosure? Or attributional style?  The inability to precisely and consistently measure such variables makes it difficult to draw strong conclusions with the research.

Much of the research done on relationships is done in Western cultures.  This is a sampling bias which may not explain why non-Western cultures stay in a relationship – or end it.  In addition, many researchers, like Hatfield, use students to study relationships.  This is problematic as students may not have the goal of a long-term relationship.  This type of sampling is not representative of a broad range of relationships. Finally, a lot of the research on the end of relationships is the result of marriage counselling.  This is also a sampling bias – studying people who recognize that their marriage is in crisis and want to do something about it.  This may mean that the sample has a certain level of education or socioeconomic status.  It also means that there are some factors that make the person want to save the marriage. More research should be done on marital satisfaction.

Why relationships fall apart is an interesting question; trying to find the answer is not easy.  Sampling biases, the correlational nature of the research and the difficulty in the operationalization of variables mean that much of the research is limited in its predictive validity. Replications of the research that show consistent findings help to strengthen the findings, but it may be there is no simple answer to this complex problem – and that relationships may fall apart as a result of a range of variables.