Factors a ec ng a rac on: Filter theory
Choosing a partner-Filter Theory (Kerckho & Davis, 1962)
These researchers compared the a tudes and personali es of student couples in short term and
long term rela onships. They devised a ‘ lter’ theory to explain how such roman c rela onships
form and develop. We start with a wide eld of ‘availables’- Everyone who is poten ally available to
start a rela onship with you. However, they might not all be appealing to us!
This, using a series of lters at various stages of the rela onship, is then narrowed down to a partner
choice from a eld of ‘desirables’
1st level of lter-Social demography
Variables such as age, social background and geographical loca on (proximity)
Such variables in uence the chances of poten al partners mee ng each other in the rst place.
The key to proximity is accessibility- our most meaningful interac ons are with people who are
nearby. It requires li le e ort to meet people who live in the same area or go to the same school or
work place as you on a regular basis.
The range of people we come into contact with is restricted due to social circumstances. We usually
“mix” with people of similar ethnic, social & educa onal groups as well as those we live close to.
Those that fall out of this category e.g. live too far away or are too middle class will be struck o the
poten al partner list!
This type of ltering leads to homogamy, meaning you are more likely to form a rela onship with
someone who is socially and culturally similar.
2nd level of lter-Similarity in a tudes
Once the eld of availables has been narrowed down by the rst lter, partners will o en share
important beliefs, values and social/cultural characteris cs.
Kerckho and Davis found that if there is a degree of agreement on a tudes and basic values then
the rela onship is likely to become more stable. Through their disclosures to each other, individuals
are able to weigh up their decisions about whether to con nue or terminate their rela onship.
Partners who are very di erent to the individual in terms of their a tudes and values are not
considered suitable for a con nuing rela onship, and so are ‘ ltered out’ from the eld of possible
long-term partners.
3rd level of lter- Complementarity of needs
It refers to how well two people t together (complement one another) and meet each other’s
needs. People who have di erent needs (e.g. the need to be caring and the need to be cared for) are
a racted to each other because they provide mutual sa sfac on. Finding a partner who
complements us ensures our needs are met. E.g. a young woman who lacks economic resources may
be a racted to an older man who can meet these needs. Not exactly “opposites a ract” – but long-
term, people are a racted to those with harmonious needs.
, Key Study – Kerckho & Davis (1962)
Kerckho and Davis carried out a longitudinal study of 94 da ng couples at Duke University in the
US. Each partner in the couple completed two ques onnaires assessing the degree to which they
shared a tudes and values and also the degree of need complementarity.
Seven months a er the ini al tes ng, the couples completed a further ques onnaire assessing how
close they felt to their partner compared to how they felt at the beginning of the study. The
researchers believed that this would indicate ‘progress toward permanence’ in the rela onship.
Main ndings
In the ini al analysis of the results, only similarity appeared to be related to partner closeness.
However, when the researchers divided the couples into short-term (those who had dated for less
than 18 months) and long-term (those who had dated for more than 18 months), a di erence
emerged. For those couples that had been seeing each other for less than eighteen months,
similarity of a tudes and values was the most signi cant predictor of how close they felt to their
partner. For those who had been da ng for more than 18 months, only complementarity of needs
was predic ve of how close each individual felt to their partner.
Di erent “levels” of the lter theory become more important as the rela onship progresses (short-
term = similarity of a tudes & values, long-term = complementarity of needs).
Levinger (1970)
Levinger (1970) a empted to replicate Kercho & Davis’ study.
330 “steadily a ached” couples went through the same procedures. No evidence that either
similarity of a tudes & values or complementarity of needs in uenced progression towards
permanence in rela onships. There was also no correla on between length of rela onship & these
variables. The validity of the ques onnaire used was also ques oned (may not have been
appropriate given the changes in social values and courtship pa erns that had occurred in the
intervening years between the two studies).
The ndings from Levinger therefore ques on the reliability and temporal validity of Kerckho &
Davis’ original results.
“Perceived” vs. actual similarity
Perceived similarity may be more importance than actual similarity – research has generally
supported the importance of a tudinal similarity in a rac on e.g. people are more likely to be
a racted to others who share many common a tudes to them rather than just a few.
However, other experiments have found that perceived similarity (ie how much we think we have in
common with someone else) is a be er predictor of roman c a rac on than actual similarity.
Choosing a partner-Filter Theory (Kerckho & Davis, 1962)
These researchers compared the a tudes and personali es of student couples in short term and
long term rela onships. They devised a ‘ lter’ theory to explain how such roman c rela onships
form and develop. We start with a wide eld of ‘availables’- Everyone who is poten ally available to
start a rela onship with you. However, they might not all be appealing to us!
This, using a series of lters at various stages of the rela onship, is then narrowed down to a partner
choice from a eld of ‘desirables’
1st level of lter-Social demography
Variables such as age, social background and geographical loca on (proximity)
Such variables in uence the chances of poten al partners mee ng each other in the rst place.
The key to proximity is accessibility- our most meaningful interac ons are with people who are
nearby. It requires li le e ort to meet people who live in the same area or go to the same school or
work place as you on a regular basis.
The range of people we come into contact with is restricted due to social circumstances. We usually
“mix” with people of similar ethnic, social & educa onal groups as well as those we live close to.
Those that fall out of this category e.g. live too far away or are too middle class will be struck o the
poten al partner list!
This type of ltering leads to homogamy, meaning you are more likely to form a rela onship with
someone who is socially and culturally similar.
2nd level of lter-Similarity in a tudes
Once the eld of availables has been narrowed down by the rst lter, partners will o en share
important beliefs, values and social/cultural characteris cs.
Kerckho and Davis found that if there is a degree of agreement on a tudes and basic values then
the rela onship is likely to become more stable. Through their disclosures to each other, individuals
are able to weigh up their decisions about whether to con nue or terminate their rela onship.
Partners who are very di erent to the individual in terms of their a tudes and values are not
considered suitable for a con nuing rela onship, and so are ‘ ltered out’ from the eld of possible
long-term partners.
3rd level of lter- Complementarity of needs
It refers to how well two people t together (complement one another) and meet each other’s
needs. People who have di erent needs (e.g. the need to be caring and the need to be cared for) are
a racted to each other because they provide mutual sa sfac on. Finding a partner who
complements us ensures our needs are met. E.g. a young woman who lacks economic resources may
be a racted to an older man who can meet these needs. Not exactly “opposites a ract” – but long-
term, people are a racted to those with harmonious needs.
, Key Study – Kerckho & Davis (1962)
Kerckho and Davis carried out a longitudinal study of 94 da ng couples at Duke University in the
US. Each partner in the couple completed two ques onnaires assessing the degree to which they
shared a tudes and values and also the degree of need complementarity.
Seven months a er the ini al tes ng, the couples completed a further ques onnaire assessing how
close they felt to their partner compared to how they felt at the beginning of the study. The
researchers believed that this would indicate ‘progress toward permanence’ in the rela onship.
Main ndings
In the ini al analysis of the results, only similarity appeared to be related to partner closeness.
However, when the researchers divided the couples into short-term (those who had dated for less
than 18 months) and long-term (those who had dated for more than 18 months), a di erence
emerged. For those couples that had been seeing each other for less than eighteen months,
similarity of a tudes and values was the most signi cant predictor of how close they felt to their
partner. For those who had been da ng for more than 18 months, only complementarity of needs
was predic ve of how close each individual felt to their partner.
Di erent “levels” of the lter theory become more important as the rela onship progresses (short-
term = similarity of a tudes & values, long-term = complementarity of needs).
Levinger (1970)
Levinger (1970) a empted to replicate Kercho & Davis’ study.
330 “steadily a ached” couples went through the same procedures. No evidence that either
similarity of a tudes & values or complementarity of needs in uenced progression towards
permanence in rela onships. There was also no correla on between length of rela onship & these
variables. The validity of the ques onnaire used was also ques oned (may not have been
appropriate given the changes in social values and courtship pa erns that had occurred in the
intervening years between the two studies).
The ndings from Levinger therefore ques on the reliability and temporal validity of Kerckho &
Davis’ original results.
“Perceived” vs. actual similarity
Perceived similarity may be more importance than actual similarity – research has generally
supported the importance of a tudinal similarity in a rac on e.g. people are more likely to be
a racted to others who share many common a tudes to them rather than just a few.
However, other experiments have found that perceived similarity (ie how much we think we have in
common with someone else) is a be er predictor of roman c a rac on than actual similarity.
