Filter Theory

Factors Affecting Attraction: Filter Theory

Kerckhoff and Davies (1962) proposed a theory of how romantic relationships form. There are factors which narrow down the ‘field of availables’ (everyone who potentially could form a relationship with a person), which are of differing levels of importance.

1st__ filter level - social demography:__ These are factors influencing the likelihood of partners meeting in the first place. They include proximity (how close the people are to each other geographically), social class, education level, occupation, religion, ethnicity, and so on. People spend more time with others who are similar to them in the above respects, meaning those who are different are less likely to become romantic partners. This leads to ‘homogamy’- forming a relationship with someone similar to someone else, socially and culturally. Having shared experiences and backgrounds is attractive to potential partners.

2nd__ filter level - similarity in attitudes:__ The field of availables has been narrowed by demography, meaning it is likely that those left share similar values, attitudes and beliefs. Kerckhoff and Davies found that similarity in attitudes seemed important in those they studied who were currently in a relationship for less than 18 months. Having the same attitudes encourages deeper communication and more self-disclosure. Couples who have little in common are less likely to last for a significant length of time.

3rd__ filter level - complementarity:__ This is the ability of partners to meet each other’s needs. A characteristic of one partner is complemented by the other, for example one partner likes to use humour, and the other enjoys being made to laugh. Kerckhoff and Davies found that complementarity was important in couples who had been together for longer in 18 months. This makes relationships work as the two partners see themselves as ‘fitting together’ well.

Filter Theory, figure 1


  • Winch (1958) found that similarity of interests, attitudes and personality are typical of the early stages of a relationship, and that complementarity of needs was more important than physical attractiveness. This lends support to filter theory.
  • There is some evidence (Anderson et al, 2003; Davis and Rusbult, 2001) that people who spend lots of time together (not just romantic partners) become more similar in beliefs and attitudes over time. This suggests that being in a relationship may cause the similarity between partners, rather than being the cause of the relationship in the first place, so weakening filter theory.
  • Levinger (1974) suggests that research has not supported the original predictions of filter theory, pointing out that just because a couple has been together longer than 18 months, that does not mean they have a greater ‘depth’ of relationship. Filter theory may not be applicable to all couples, and may be even less valid in other cultures.

Social Exchange Theory

Social exchange theory is an example of an ‘economic theory’ of a relationship. Such theories suggest that people in relationships wish to give and receive something, assuming that people act out of self-interest ultimately.

Thibault and Kelley (1959) suggest that relationship satisfaction is judged by considering the ‘rewards’ and ‘costs’ the relationship produces. People seek to minimise the losses and maximise the rewards, leading to ‘profit’ (the ‘minimax principle’). What counts as rewards and costs will vary from person to person and relationship to relationship, and will change over time. They include sex, emotional support, praise (rewards) and stress, energy, having to compromise (costs). The ‘opportunity cost’ refers to the investment in a current relationship preventing a person investing resources elsewhere.

Comparison level (CL): This is one measure of the profit in a relationship, referring to the amount of reward a person thinks they deserve to get. It is influenced by previous relationship and social norms (affected by the media). If the person’s CL is high, they will think the relationship is worth pursuing. People with high self-esteem are more likely to have a high CL (and low CL for those with low self-esteem) meaning the expectation of reward is greater.

Filter Theory, figure 1

Comparison level for alternatives (CLalt): The second measure of profit. This involves a person considering whether they would get more rewards, and fewer costs, from another relationship, or from being single. According to the theory, the person will remain in the relationship if they think that the current relationship will give greater profit than the alternatives. This is affected by the state of the current relationship- if it is satisfying, there is much less need to consider alternatives.

Stages of relationship development: Thibault and Kelley suggest each relationship develops through four stages:

  • Sampling stage: exploring the rewards and costs of relationships (not just romantic relationships) through experimentation
  • Bargaining stage: the beginning of a relationship, where various costs and rewards are identified through exchange and negotiation
  • Commitment stage: costs and rewards become more predictable and the relationship becomes more stable- rewards increase, costs lessen
  • Institutionalisation stage: the norms of the relationship are firmly established


  • It has been argued that economic exchange characterises non-romantic relationships (such as between work colleagues), but do not characterise romantic relationships, as neither partner keeps ‘score’ of who is ahead and behind on rewards and costs. SET may be based on a faulty assumption and is therefore weakened.
  • Miller (1997) found that those who rated themselves as being in a committed relationship spent less time looking at images of attractive people, suggesting that dissatisfaction in a relationship is what leads to the costs/rewards being calculated, rather than the calculation leading to dissatisfaction. This weakens the prediction of SET.
  • There is evidence that measures of fairness are more important in relationships than calculation of costs and rewards, suggesting that SET is, at best, a limited explanation of romantic relationships.

Equity Theory

Equity: This refers to fairness. Walster et al (1978) suggested that what is most important is that the level of ‘profit’ in a relationship is roughly equal for both partners. If it is not, then one partner overbenefits, and the other underbenefits, which may lead to resentment and anger in the ‘underbenefitted’ partner, so threatening the relationship. If the relationship is perceived to be fair, both partners will be satisfied.

The amount of rewards and costs are not thought to matter according to this theory. What matters is the ratio between them- investing a lot in a relationship is acceptable, as long as the level of rewards is high. Satisfying relationships are characterised by negotiations to ensure equity, but not necessarily equality, amongst the distribution of rewards.

Consequences of inequity: If a partner invests a lot in a relationship, but gets little out of it, then they will become dissatisfied. There will be a correlation between the level of perceived inequity and the level of dissatisfaction. Both the overbenefitted and underbenefitted partner will notice the inequity. The perception of inequity can change over time, for example, contributing more than what is received may be acceptable early on in a relationship, but will be perceived as unfair if it continues for a long period of time. To deal with the inequity, a partner may work harder to try to restore equity. Alternatively, a partner may cognitively revise their perceptions of what counts as rewards and costs, so that the relationship comes to be seen by them as equitable, even though nothing has really changed.


  • Utne et al (1984) found that couples who considered their relationship equitable were also more satisfied than those who reported themselves as underbenfitting or overbenefitting, so supporting the predictions of equity theory.
  • Aumer-Ryan et al (2007) found cultural differences in the link between equity and satisfaction. Those in collectivist cultures were more likely to be satisfied when overbenefitting in a relationship, whereas in individualist cultures equity was more associated with satisfaction. This suggests equity theory may be less applicable in different cultures.
  • Huseman et al (1987) argue that not all people are concerned with the need for equity. Some take more satisfaction from contributing more to a relationship, whereas others are prepared to ‘overbenefit’ without guilt. This weakens the theory, as it does not account for individual differences.

Rusbult’s Investment Model

Rusbult et al (2011) suggested that commitment is an important factor in relationships, referring to the intention or desire to continue the relationship (believing it to have a future). Commitment is affected by the following factors.

Filter Theory, figure 1

Satisfaction and comparison with alternatives: This is similar to the ideas of social exchange theory. In relationships, each partner considers the rewards and costs (comparison level- CL), and satisfaction will be high if there is seen to be profit from the relationship. If a partner does not consider they will get more profit elsewhere, the relationship is more likely to continue. However, satisfaction by itself will not be enough to determine the continuation of a relationship.

Investment size: Investment refers to the extent and importance of the resources associated with the relationship. If the relationship ends, this investment is lost. There are two types. Intrinsic investments are resources such as money and possessions which are put into a relationship by the individual partners. They also include things like energy, emotion, and self-disclosures. Extrinsic investments are factors brought by the relationship, such as a house, a car, mutual friends, and children. They also include shared memories.

If the size of the investment increases, along with the sense of satisfaction being seen as acceptable, each partner’s commitment to the relationship will be stronger.

Satisfaction v commitment: According to the theory, commitment is more important than satisfaction. This can explain why people who are dissatisfied still continue with a relationship. It is because they do not want the investments they have put in to go to waste, so they will work hard to try to increase satisfaction and repair the relationship.

Relationship maintenance mechanisms: Each partner puts in work to promote and maintain a relationship, and will put their partner’s interest first, forgiving any serious transgressions from the partner. Cognitive strategies are also used to maintain a relationship, for example, thinking in an unrealistically positive way about the partner, and being negative about other people’s relationships (so making a negative comparison with alternatives).

Evaluation of Rusbult’s Investment Model:

  • Le and Agnew (2003) found that in 52 studies with 11,000 participants, satisfaction, comparison with alternatives and investment size all predicted relationship commitment. Where commitment was greatest, relationships were longer-lasting. This was true across different cultures and in homosexual as well as heterosexual relationships, strongly supporting the theory.
  • The theory can be used to explain why partners stay in abusive relationships, where the abused partner clearly cannot be satisfied. There is evidence that female victims of domestic abuse who stayed with their partner were more likely to report having invested a great deal in the relationship, supporting the prediction of the theory.
  • Investment may have been oversimplified in the theory. This is because in the early stages of a relationship, little investment is made by either partner, but this does not mean the relationship does not last. Therefore, investment may need to be extended to include factors such as future planning. This means the original explanation may be incomplete.

Duck’s Phase Model

Duck (2007) proposed an explanation of why relationships break down. He argued that the end of a relationship is not a sudden event, but a process which goes through phases where the perception of the relationship changes once a ‘threshold’ is reached.

Intra-psychic phase: (‘I’m not satisfied/I can’t stand this anymore’) The dissatisfied partner begins to think more negatively about the relationship, focusing on why they are dissatisfied and their partner’s shortcomings. They weigh up the pros and cons of the relationship continuing and may confide their feelings in a friend.

Dyadic phase: (‘I would be justified in ending the relationship’) Here, the couple start to talk to each other about the relationship, in a confrontational way. Discussions take place over the lack of fairness and a rethinking of the commitment to the relationship. Often discussions will be quite hostile. Either the couple makes a renewed attempt to save the relationship, or will be determined to continue breaking it up.

Filter Theory, figure 1

Social phase: (‘This is over- I mean it’) In this stage, friends and family become aware of the problems, as the break-up is made public. Each partner will try to gain the support of particular friends, forming pacts in which they are supported. The blame may be attributed to one partner by the couple’s social networks, and people may contribute information which hastens the break-up, for example a secret that one of the couple have been keeping. Alternatively, efforts may be made to salvage the relationship. Usually however, a relationship will not recover once this stage is reached.

Grave-dressing phase: (‘The break-up is inevitable‘) This refers to the aftermath of the break-up, where each partner creates a favourable story about why the relationship ended, for example by blaming the other partner or circumstances. Gossip is important, as each partner will try to retain ‘social credit’ by not putting the blame on themselves. Cognitive techniques are used during this phase to ‘rewrite’ the history of the relationship, for example by interpreting characteristics previously seen as positive or neutral as negative. Alternatively, both partners may split more amicably, agreeing to move on with their lives.


  • The original model has been criticised as not accounting for the fluid and dynamic nature of relationship breakdown. Rollie and Duck (2006) attempted to address this by adding an explanation that progression from one stage to the next is not inevitable, or linear. They also added a fifth stage, the ‘resurrection phase’, in which ex-partners use the lessons from the ended relationship in a future relationship.
  • It is hard to investigate this model empirically, especially the early stages, as researcher involvement may make a break-up more likely (as partners may be encouraged to think and speak about their grievances). Much of the data is retrospective and based on recalling information once the relationship is over. This is likely to be less valid and reliable.
  • The model has been accused of cultural bias. In collectivist cultures, there may be more of a sense of obligation in the relationship, so it is less likely to end. In addition, friends and family may be more involved in the process. This means Duck’s model may not be applicable to all cultures.
Describe and evaluate Duck’s phase model of relationship breakdown. (16 marks - 6 outline - around 3 paragraphs; 10 evaluate - around 3-4 evaluation points)
Your answer should include: Intra-physic / Dyadic / Social / Grave-dressing / Simplistic / Methodical / Problems / Cultural / Bias