Such attraction may lead to the formation of a relationship. This is likely to differ, as with heterosexual relationships the needs of a man and a woman are likely to be different, whereas if two people of the same gender form a relationship, the needs are likely to be similar. This may lead to problems in initiating activity, or may cause competition, a viewpoint supported by Kitzunger & Coyle, who found that in the formation of lesbian relationships, it was necessary for one partner to be more assertive than the other in order for dates, intimacy, etc., to be initiated.
The maintenance of homosexual relationships may differ to some extent to that of heterosexual relationships for similar reasons. Whereas traditionally in heterosexual relationships the man and woman differ in terms of rewards seeked and efforts needed to be put in, in homosexual relationships these may be more similar. Indeed, Huston & Schwartz found that cohabiting homosexual men had more sex on average than any other type of cohabiting couple, whereas lesbians had the least. This is reflected by the sociobiological theory's depiction of men as 'spreading their seed' and women as 'tending and befriending' others. As a result, maintenance strategies may differ from homosexual relationships. Huston & Schwartz found that in gay male relationships, confrontational conversation was used as a means of 'sparring' for a higher status, whereas in lesbian couples it was avoided as it would pose a risk to intimacy and cooperation. This is however contested by Kollock et al., who found that the more assertive partner in a lesbian relationship would often use aggressive conversational tactics such as interruption.
Homosexual and heterosexual couples may also differ in terms of breakdown. An example of this is the finding by Becker that lesbian relationships often break down when one partner falls in love with another woman, and that the breakup is initiated by that partner. However, another study found that infidelity in homosexual relationships does not necessarily cause disagreement between the partners.
The above research accentuates the differences between homosexual and heterosexual relationships, but it is much more likely that they are more similar than different. Silverstein argues that this difference arises because homosexual samples are usually derived from gay bars and gay neighbuorhoods, which are characterised by temporary and casual, often sexual, relationships. It is likely that the results therefore paint a negative image of homosexual relationships, and that the sample and findings are unrepresentative of the general homosexual population.
Computer-mediated (CM) relationships are another understudied type of relationship. This relatively recent phenomenon may have occurred as a result of the increased popularity of the internet, and is when relationships form or are maintained without the necessity of geographical proximity. According to Hellerstein, 'heavy' users of e-mail were likely to form more relationships online, many of which then continued 'offline', whereas light users of email formed relationships in other ways. Chenault argues that such relationships may be successful because we seek to read between the lines of text-based speech; that is, small differences such as punctuation, capitalisation and the use of 'smileys' are likely to greatly affect the interpretation of the text. Therefore, we may derive as much socially from CM conversation as from face-to-face covnersation. This argument is supported by Walther, who founded that participants reported perceiving slow replies to email as being more intimate than quick responses. This indicates a degree of 'reading between the lines'. However, other researchers have argued that CM relationships cause a 'social vacuum' which cannot satisfy a person's needs with regard to what can be derived from a relationship. They argue that the 'deindividuation' that occurs by means of CM relationships makes it impossible to know the true personality of the other person in the relationship. However, Young argues that this 'deindividuation' is beneficial for those who have low self-esteem or are too shy to initiate a relationship face-to-face, and may allow them to express their feelings in a more secure way before continuing the relationship.