Back to Journal Cover Page


Volume 6, Number 5
Submitted: November 29, 2000
Resubmitted: February 8, 2001
Accepted: February 13, 2001
Publication date: February 16, 2001


Carlos Yela
University of Madrid

Jose Luis Sangrador
University of Madrid


We try to verify various hypotheses about the importance of physical attractiveness (PA) in loving relationships, based on known psycho-social processes. We select a representative sample of the Spanish population (n = 1949) and evaluate the PA subjectively perceived by subjects in themselves and in their partners. This evaluation is carried out using self-report techniques, measuring PA both at the present moment and retrospectively at the start of the relationship. The results suggest that we tend to form loving relationships with people perceived as similar in attractiveness to ourselves (matching hypothesis). We also tend to perceive our partner, in a biased way, as a little more attractive than ourselves. Furthermore, with the passing of time, the effects that would tend to diminish the perceived PA of our partner (habituation, differential reinforcements value, etc.) seem to be counterbalanced by others which tend to increase it (familiarity, cognitive dissonance, etc.).



Whilst little recognized (by people in general) as an important factor in people’s choice of partner (Hadjistavropoulos & Genest 1994), social psychology has made clear the important role played by physical attractiveness (PA henceforth) in our social cognition (e.g. Metee and Aronson 1974; Sangrador 1993), behaviors (e.g. Griffitt 1979), and interpersonal relationships (especially in loving relationships; e.g. Cook & Wilson 1979).

In a recent study (Sangrador and Yela 2000), and in line with the consequences of the well-known "halo effect" " (Dion et al. 1972; Feingold 1992), we showed the influence of the PA of the loved one on the desire to initiate a relationship with them, the manner of falling in love, the intensity of feelings of love, idealization of the other, and even satisfaction with the relationship (see also Yela 2000a). In the current study, our aim is to investigate the possible relationship between the PA of both partners in a couple, and the course that each one’s PA takes throughout the duration of the relationship. To our knowledge, empirical studies of this precise point are very scarce (f.ex. Margolin & White 1987; Zajonc 1987)

At the same time, it is important to stress certain refinements which characterize the present study and differentiate it from common studies in this area:

First of all, the dimension of PA has more often been investigated in studies of attraction than of love. Thus, for example, the effect of PA in first interactions, or in the formation of an impression of the other is studied (from the classic studies of Byrne 1971 to more recent studies by Chen et al. 1997). Research dealing with the role of PA in loving behavior however, as in the present investigation, are few and far between.

Secondly, a significant proportion of research has endeavored to operationalize the variable PA in an "objective" manner (for example by means of an "objective" evaluation – in reality inter-subjective – of PA by a panel of "judges"). In the present investigation however, PA is, strictly speaking, "perceived" PA (PPA henceforth), perceived by the interviewee both in themselves and in their partner, and so it may be independent of both partners’ "objective" PA, if such thing could exist.

Thirdly, while the majority of the literature in this area normally speaks of PA restricted to the PA of one person, and in general a stranger, we have considered the interviewee’s PA for him or herself (PPAS henceforth), as well as the interviewee’s PPA off his or her partner (PPAO henceforth, the "O" representing the "other"). Furthermore, we have considered these dimensions both at the present moment and at the start of the relationship (in this last case indicated by a subscript "o", for origin). This allows very subtle analysis from diverse perspectives.


Lastly, while the immense majority of empirical studies are carried out with samples of students (or in any case with non-representative samples), our data have been collected from a representative sample of the Spanish population (see also Barrón et al. 1999).

In the current study, we aim to verify what is often referred to in classic positivist language as a confirmatory hypothesis, a critical hypothesis, and an exploratory hypothesis, based on different processes and classical theoretical viewpoints in social psychology:

Confirmatory Hypothesis
The well-known matching hypothesis predicts that (even though we desire people with the highest possible PA) our partners are in fact people with a PA similar to our own (see, for example, Murstein 1972, and some controversial papers: Kalick & Hamilton 1986; Aron 1988). We attempt to verify, on a non-student and non-US representative sample, a subjective version of this hypothesis: that we perceive our partner to be about as attractive as we perceive ourselves to be.

So on one hand, we tend to feel attracted to, and to desire people with a higher PA. This is the case across a range of diverse human cultures, regardless of differences in aesthetic criteria (Wilson & Nias 1976; Cook & McHenry 1978, Feingold 1990, 1992; Buss 1994, 1998; Sangrador & Yela 2000). But on the other hand, at the same time it is true that our partners tend to be people similar to ourselves in terms of PA. This is known as the matching hypothesis, although it has been repeatedly verified (Murstein 1972; Huston and Levinger 1978; Feingold 1990; Wong et al. 1991, etc.). This has been explained, within exchange theory, using the concepts of a balance between costs and rewards, and a search for equity (Walster et al. 1978; Critelli & Waid 1980). In this context, PA is a powerful personal reinforcement (by means of socialization in the relationship "beautiful-good"; Sangrador 1993) and a powerful social reinforcement, due to the high importance socially attached to this characteristic, even though it is not often acknowledged. A partner much more attractive than oneself could involve a series of additional costs and additional effort in both starting and maintaining the relationship. Similarly, a much less attractive partner involves a high cost (or a deficit in rewards, at least in physical appearance), for which reason either the partner is not chosen, or else is chosen at the cost of an increase in the PPAO, as suggested by the cognitive dissonance perspective (Festinger 1957). Furthermore, similarity in general has been proved as one of the main factors in the courtship, mate preference and mate selection processes (since Reiss 1960; Byrne, 1971; Lewis 1972).


Critical Hypothesis
If the effects of habituation and of relative value of reinforcements (Skinner 1953; Aronson & Linder 1965; Frijda 1988) prevail, then the PPAO will decay with time, but if the effects of familiarity and cognitive dissonance prevail (Festinger 1957; Aronson and Mills 1959; Zajonc 1968) then the PPAO will be increased throughout the relationship.

As is well-known, one of the most important factors relative to all psycho-social processes, and perhaps in particular to the one that interests us (i.e., the loving relationshi), is the passing of time (in fact, for that reason we speak of processes instead of structures). It seems evident that we do not feel (nor think, nor perceive) the same feeling for a stranger who we are suddenly attracted to, as we do for a new lover who we are getting to know, or for a person with whom we have shared the last, say, ten thousand days (and the last ten thousand nights, one could cynically add). It should be noted in any case, that to affirm that one doesn't feel the same does not imply any value judgement about whether one feels more or less, or if they are better or worse feelings). How does that passing of time affect in our specific example the perception of the partner's PA?

On one hand, given that repeated exposure to a stimulus tends to reduce the intensity of the elicited response (according to the well-known law of habituation; Skinner 1953), we might suppose that living together with the passing of time can diminish the response of physical attraction that our partner provokes. In a similar way, according to the gain-loss theory (Aronson and Linder 1965), the subjective value of reinforcements is not measured in absolute terms but rather is relative to our reinforcement history. In this way, the power of reinforcement of our partner's physical attractiveness would tend to reduce in intensity as we get used to him or her. Finally, according to the change of emotions law, and the comparative feeling law (Frijda 1988), which indicate that the hedonic value of a stimulus depends on the baseline and on the change in the situation, more than on the stimulus or situation itself, we could make an identical prediction: the perception of the partner's PA will tend to decrease with time.

There are many other processes however, that would imply the opposite. For example, the effects of familiarity and mere exposure (Zajonc 1968) explain the attraction that we experience toward stimuli that are well-known (familiar), and even the attraction raised by the mere exposure to a stimulus (without positive hedonic value). In this way, on many occasions we find people physically more attractive as we get to know them (and, therefore, expose ourselves to them and make them familiar) than we do just after meeting them. Something similar could happen within the couple. And on the other hand, as a corollary of cognitive dissonance theory (Festinger 1957), it is known that we tend to value more a posteriori, that which it is supposed to have been a personal and free choice, and which requires of us an effort to obtain and/or to maintain (Aronson and Mills 1959). If we value PA (and we do so greatly, as we have shown in a separate study; Sangrador and Yela 2000) and we perceive that our partner (who we have chosen freely and who requires of us an effort "x" to maintain) no longer seems as attractive as before, a state of cognitive dissonance is produced, which will tend to be reduced, by either reducing the value assigned to PA (something easy to do verbally but not so easy internally, given the socialization process and opposing social pressure), or by convincing ourselves that he/she continues to seem as attractive as on the first day (or even more).


While recognizing that other contingencies exist which can affect the perception of the partner's PA over time (possible negligence in care for physical appearance with time, an eventual "objective" increase in physical attractiveness, etc), we will try to establish which of the aforementioned processes is more likely to predominate according the empirical fluctuation of PPAO.

Exploratory Hypothesis: If the matching effect is verified, will that similarity of perceived PA remain throughout the relationship, or will it decay with the passing of time?

Will the matching effect prevail or will biased information processing prevail, so that the partner is perceived as more and more attractive than oneself? Can both effects be compatible, so that the PA perceived in the partner and in oneself will be relatively similar throughout the relationship, while always superior that perceived in the partner to that attributed to oneself? If it is the case, it could be that both the matching effect and certain cognitive processes (cognitive dissonance, social desirability, re-establishment of equity, etc.) intervene when formulating the perception of our partner's PA, and our own PA, throughout the relationship.

We will try to find answers to these questions in the third part of the study.


This study falls within the framework of a larger investigation into loving behavior conducted by means of individual interviews with a representative sample of the Spanish population. Here we refer only to those methodological aspects related to the testing of the aforementioned hypotheses.

The sample was composed of 1,949 individuals of both sexes, between 18 and 64 years of age, resident in towns with a population of over 2,000 inhabitants throughout Spanish territory. The sampling procedure was multi-stage, stratified by conglomerates, with selection of the primary units (towns) and the secondary units (sections) carried out in a random proportional manner, and of the last units (individuals) by random means considering sex and age quotas, so that the definitive sample is highly representative of the Spanish population (sampling error ± 2.23, with a level of confidence of 95.5%).


Variables and Measurement Technique
Four basic variables of PA have been used, each of them assessed by means of the interviewee’s own self-report, with an item range 0 (minimum PA) to 10 (maximum PA). Two of the variables refer to the perceived PA (PPA) for the interviewee in his/her partner or "other" (PPAO), both at the beginning of the relationship and at the present moment. The other two refer to the perceived PA by the interviewee in him or herself (PPAS), both at the beginning of the relationship and at the present moment. Naturally, the questions about the beginning of the relationship are of retrospective character, with all the limitations that this implies (basically referred to the possibility of perceiving the past biased by the present situation: in our case, the possibility to perceive PPAOo biased by the PPAO tending to adjust them, and even biased by the present level of satisfaction with the partner and with the relationship).

Therefore, the four basic variables are configured as follows:

For the first three variables data has only been obtained from the interviewees with partners (involved in a loving relationship) at the moment in which the interview was carried out. For the last variable data has been obtained from all subjects, both those with partners and those without. Obviously, we have only taken into account the PPAS data from subjects with partners when carrying out the comparative analyses between the PA of both members of a couple.


Lastly, we introduced two further variables operating with the four basic variables:

In addition, information was gathered about the variable "duration of the relationship" and "cohabitation" (whether the couple lives together or not), by asking the subjects directly.

Procedure and Analysis Technique
Each interview was carried out individually in the interviewee’s own home, and anonymity was guaranteed to all subjects. Interviewees were selected according to the sampling procedure detailed above. Once the responses had been coded, different types of data analysis were carried out; in the current study descriptive statistics, linear correlations, variance analysis, difference of averages and regression analysis are included. The statistical package used was SPSS 9.0 for Windows.


Confirmatory Hypothesis: The Matching Effect
The first part of the hypothesis, that we love people with higher PA, is not really tested in this study, but it has been repeatedly verified (e.g. Wilson and Nias 1976; Cook and McHenry 1978; or more recently Buss 1998). In a recent study we verified that PPAO was the characteristic most valued in the partner (out of a long and varied list of characteristics) for short-term relationships, and that it also had an important effect in long-term relationships: on falling in love, the intensity of loving feelings, idealization and personal satisfaction with the relationship (Sangrador and Yela 2000).

Let us see now if our hypothesis (i.e., in spite of the above, we tend to form relationships with people of a similar PA to our own, or more exactly, with people of a similar PPA to our own perceived one) is verified. If this were not true, we could expect a random distribution of PA so that the correlation between partners PA was not significantly different from zero.


Pearson’s correlation coefficient between PPASo and PPAOo is .43 (p <.000; n = 1078). This is the appropriate index to verify the matching hypothesis, since in this case we are not so much interested in the correlation between current values (that is also relatively high and significant; r = .39; p <.000; n = 1251), as in the correlation between values at the origin of the relationship. The fact that the correlation obtained is significantly different from 0 indicates the confirmation of the hypothesis. The fact that it is different from 1 indicates that the correspondence is not perfect, absolute and certain; but, as is almost universally the case in psychology, a trend. That is to say, it is not that we always form relationships with people of PA exactly the same as our own, but rather that we tend to form relationships with people of a similar PA to our own, or at least who we perceive as similarly attractive to ourselves.

In any case, and while it is necessary to note that in this case the difference between averages is not the best statistic (because it can hide intra-couple differences and compensate them with inter-couple differences of similar magnitude and opposed direction). We should also point out that we obtained a significant difference between the average value of the perceived PA in the other and in the self at the start of the relationship, the values obtained being PPAOo = 8.55 (SD = 1.71) and PPASo = 7.41 (SD = 2.09) (p <.000; n = 1078). This seems to indicate that, at the start of the relationship, besides the matching effect, a favorable perceptual bias tends to take place towards the partner with respect to the perception of PA; a bias that may be explained on different ways, for example in terms of enhancing self-esteem, or (as it was suggested by a reviewer) simply as normative modesty.

Critical Hypothesis: PPAO and the Passing of Time
Initially the value of the linear correlation between the PA perceived in the partner and the duration of the relationship is not significantly different from 0 (r PPAO-duration of the relationship = .06 ns). This indicates that on average, there is no linear uniform change in PA over time, neither an increase, nor a decrease. It could be that the relationship between PPAO and the passing of time is not linear, but rather fluctuates as a function of the passing of time in a certain way. To check this possibility, we established certain bands for the variable, duration of the relationship, and plotted the curve for PPAO bearing in mind the averages of the variable PPAO in each one of the time bands. In figure 1 we can see this fluctuation.


Figure 1:

The graph raises diverse questions, although we should note that the magnitude of the fluctuations is remarkably low (in fact, an ANOVA shows significant differences only between the lowest point (7.89) and highest point (8.22) of the graph, although this depends on the selected bands). In our opinion, the pertinent point here is that we can appreciate different stages in the fluctuation of PPAO that could correspond to the preponderance of some bio-psycho-social or other processes.

Thus, the initial increase corresponds to the falling in love phase (see Yela 1998). After this, at a certain point (which varies for different individuals and couples, and perhaps between cultures) the PPAO begins to decay, as a result of the passionate components of love (Sternberg 1986; 1988; Fisher 1992; Yela 1996, 1998, 2000b). The processes mentioned in our hypothesis probably intervene here: habituation and the relative value of reinforcements (Skinner 1953; Aronson & Linder 1965; Frijda 1988 etc.).


There is however, also a point (which in our study is located in the uncertain interval between 5 and 15 years, and which should also show clear interpersonal variability), where these processes begin to be compensated by others of an opposed nature, and the PPAO increases again. The processes in question probably include those processes proposed in our hypothesis, namely the effects of familiarity and cognitive dissonance (Festinger 1957; Aronson and Mills 1959; Zajonc 1968), although another explanation could be offered for the moderate increase in PPAO during advanced stages of the relationship. We know that perception of the PA is not based exclusively on physical factors (Sangrador 1993). So it is possible that in older people PA becomes increasingly related to fundamentally psychological components, and less to specifically physical components (this distinction should be understood without the implication of any kind of ontogenetic dualism). This could attenuate the decline in the PPAO or even increase the PPAO, as a result of greater intimacy and commitment, characteristic of the late stages of loving relationships (of those relationships which survive, which are those that compose the sample) (Yela 1998). In any case, this latter explanation is perfectly compatible with the proposals in our hypothesis. (As always, there is another possible explanation. Perhaps this dip-increase pattern could be related to marital satisfaction rates and divorce trends. Dissatisfying marriages are more likely to end by 5-15 years, so that the later time periods include a higher percentage of satisfying relationships. The earlier periods may include couples on their way to divorce, who perceive their partners as being less attractive).

Finally, the small decrease that is observed in the last part of the graph suggests that the tension between those opposed forces continues to operate but no one force definitively overcomes the other.

Given the results obtained we could suggest that at first the effects of habituation prevail but later on they are compensated by those of familiarity and dissonance.

And if we keep in mind the average initial value of PPAOo (8.55), those two successive periods of habituation and familiarity-dissonance appear even more clearly. If we consider the value of PPAOo (retrospectively evaluated) we must agree that PPAO decays with time (from 8.57 to 8.06; p <.000, as shown in Chart 1), although slightly, and in any case far less than we would expect if the forces of habituation acted alone. And if we only consider the fluctuation of PPAO from the average value of the first time band (the relationships between 1 day and 1 year of duration), then we cannot even affirm that the PPAO decays with the passing of time, but rather stays stable on average, although it fluctuates in the way that we have commented.


We wonder then, how the difference perceived between the partner’s PA would fluctuate between the beginning of the relationship and the current moment. In this case the correlation obtained is significant, although quite low: r (PPAOo - PPAO) - duration = .13 (p <.000; n = 1131).

This indicates that on average for the sample, the more time passes the greater the difference between how the partner is perceived physically and how they were perceived at the beginning of the relationship (or strictly speaking, how they say now that they perceived themselves then). This means that with time, either the partner's PA at the outset of the relationship is idealized, or the perceived PA diminishes (or both things), albeit slightly.

Lastly, we try to discover the fluctuation of that difference (PPAOo - PPAO, that is to say, the fall in PPAO from the start of the relationship to the current moment) this time in function not of the passing of time, but of a variable that normally co-varies with it: living together as a couple. Variance analysis gives the following results:

Table 1. PPAOo - PPAO (PA of the other at the origin less current PA of the other)

They live together all the time. 0.55    
They live together sporadically. 0.12 F = 4.72 p < .000
They do not live together. -0.07    

As we can see, the decrease in the PA perceived in the partner from the beginning of the relationship to the present time is greater for those who live together than for those who do not. This would lean towards the explanation offered by the habituation process. The reduction observed in the PPAO should not be attributed so much to the mere passing of time itself, but to the fact that this usually implies a continuous exposure to the partner (since the vast majority - 83% - of the couples live together). The fact that the differences are not bigger could be explained by the action of the aforementioned opposing mechanisms (principally familiarity and dissonance).

Exploratory Hypothesis: Relationship between PPAO and PPAS Throughout the Loving Relationship
Again, the data gathered does not allow us to clarify this relationship definitively. Rather it seems as if, as in the previous case, that both processes (matching and cognitive biases) play a part.

As in the confirmatory hypothesis about matching, the differences in averages do not interest us (this could hide the intra-couple variability, which is what affirms matching) but rather the correlations between PPA in oneself and one’s partner.


Firstly, if the correlation between the PA perceived in the partner and the self at the start of the relationship, was of .43 (p < .000), as we saw above, the same correlation referring to the course of the relationship (not to the outset) is: r PPAO-PPAS = .39 (p <.000; n = 1251), which suggests that the matching effect tends to last over time.

On the other hand, the difference between PPAO and PPAS neither increases nor diminishes systematically throughout the relationship, at least on average: r (PPAO less PPAS) - duration of relationship = - .02 (n.s.).

At the same time however, we can appreciate that the PA attributed to oneself, at least on average, is always significantly smaller (around 1 or 1.5 points out of 10) than the PA attributed to the partner. In the verification of the matching hypothesis we show this for the beginning of the relationship. The same effect can be observed for relationships in progress, PPAO being = 8.04 (SD = 1.79) and PPAS = 6,49 (SD = 2.10), also a significant difference (p < .000; n = 1251). Such a result suggests the influence of the aforementioned perceptual biases (relative over-valuation of the PA of the other)

Finally, let us see what is the correlation between the PA perceived in oneself and in the partner (that is to say, matching) at the different temporal points that we have defined in the variable, duration of the relationship (taking p < .000):

Table 2.

r PPAOo-PPASo 0.43
r PPAO-PPAS (for t < 1 year) 0.21 (n.s.)
r PPAO-PPAS (for t = 1 to 3 years) 0.26 (n.s.)
r PPAO-PPAS (for t = 3 to 5 years) 0.13 (n.s.)
r PPAO-PPAS (for t = 5 to 15 years) 0.40
r PPAO-PPAS (for t = 15 to 25 years) 0.42
r PPAO-PPAS (for t = 25 to 45 years) 0.43
r PPAO-PPAS (for t > 45 years) 0.46


The results suggest that the matching hypothesis is verified both at the outset of the relationship (that is to say, at the time of forming a relationship with the partner) and later, from a certain point in time (in our study, starting at around 5 years) until the end of the relationship. However, there is one period where a systematic and significant relationship doesn't seem to exist between the perception of PA in oneself and in the partner: between the beginning of the relationship and more or less, 5 years. This is perhaps not so surprising if we bear in mind that those years are in fact those that the specialized literature has denominated the falling in love and passionate love phase (Yela 1998; 2000 b). In this phase one of the most relevant processes is certainly idealization of the partner, which includes the idealization of the PA. If during this period the PPAO is high, it is plausible to assume that it will become more different from the PA attributed to oneself. On the other hand, with the diminishing of passionate factors (among them idealization), PPAO will descend (as can also be observed at the outset in figure 1), tending towards the PPAS and strengthening the relationship between them.

Finally, if the correlation stays high until the end and, as we have observed, PPAO tends to increase in late stages of the relationship (figure 1), this should mean that the PPAS does likewise. In fact this is exactly what happens.

Figure 2.


As we can see, the course of PPAS is relatively similar to that of PPAO. The later, and in principle surprising phase of growth could be attributed to any combination of a number of factors of a very diverse nature. Firstly an attempt (be it primarily conscious or unconscious) to maintain equity by means of matching, secondly an attempt to protect self-esteem (now that physical appearance in fact begins to deteriorate gradually), thirdly a conception of PA which focuses increasingly on psychological rather than specific, exclusively physical aspects (as we suggest above). Finally, a hypothetically smaller capacity for critical analysis in the sub-sample whose relationship has lasted more than 25 years, who are different in that they have a lower cultural level (lower critical analysis which could lead them to directly assign themselves the maximum PA score, as in fact is the case in more than 40% of these sub-groups).

All things considered, given the results, it seems that throughout the relationship the matching effect cohabits with the relative over-valuation of our partner's PA.

Therefore, from the results obtained we could conclude:

Nevertheless and before concluding, we should remind ourselves of the methodological limitations that, as with all empirical studies, limit our capacity to generalize from the results:


Firstly, we should remember that whenever the passing of time is an important variable in a study, the ideal situation would be to have a longitudinal design, instead of transverse as in our case: to see each couple's evolution over time, instead of comparing different couples at different moments in time, albeit on average. Secondly, the use of the self-report technique, whilst allowing us to evaluate beliefs, attitudes and feelings difficult to access using observation, always leaves us open to biases in the answers (just to mention one of them, one of the reviewers wondered whether matching hypothesis could be debt to a tendency for people who rate themselves high in PA to rate all others high in PA as well. It could be). And finally, the retrospective nature of the perception of PA at the beginning of the relationship also calls for caution when interpreting the results (as Machado wrote: "in man neither tomorrow nor yesterday is written").

Secondly, we are aware about some potential sex differences in the way men and women perceive PA throughout loving relationship. We take them into account on our next project that is already in preparation.

Nevertheless, and given that we have just highlighted the methodological limitations, it is also fair to say that the use of a representative sample of the Spanish population, besides being extraordinarily uncommon in social psychology (and in psychology in general), leaves considerable room for generalization from the obtained results. What is truly important is that these results could be replicated and confirmed in later studies. This would signify a substantial step forward in the clarification of the paradox surrounding the decisive importance which something so seemingly frivolous as PA has on something that we do not dare to consider frivolous (i.e., love and relationships between couples) (Sangrador and Yela 2000).


Aron, A. (1988). "The matching hypothesis reconsidered again: comment on Kalick and Hamilton." Journal of Personality and Social Psychology, 54:441-446.

Aronson, E. and D. Linder (1965). "Gain and loss of esteem as determinants of interpersonal attractiveness." Journal of Experimental Social Psychology, 1:156-171.

Aronson, E. and J. Mills (1959). "The effect of severity of initiation on liking for a group." Journal of Abnormal and Social Psychology, 59:177-181.


Barron, A., D. Martinez-Inigo, P. D. E. Paul, and C. Yela (1999). "Romantic beliefs and myths in Spain." The Spanish Journal of Psychology, 2(1):64-73.

Buss, D. M. (1994). The Evolution of Desire: strategies on human mating. NY. Basic Books.

Buss, D. M. (1998). "The Psychology of human mate selection." In C.B.Crawford et al. (eds.): Handbook of Evolutionary Psychology: ideas, issues and applications. pp. 405-429. Mahwach. N J. LEA.

Byrne, D. (1971). The attraction paradigm. New York: Academic Press.

Chen, N. Y., D. R. Shaffer, and C. Wu (1997). "On physical attractiveness stereotyping in Taiwan: A revised sociocultural perspective." Journal of Social Psychology, 137(1):117-124.

Critelli, J. W. and L. R. Waid (1980). "Physical attractiveness, romantic love and equity restoration in dating relationships." Journal of Personality Assessment, 44.

Cook, M. and R. McHenry (1978). Sexual attraction. Pergamon Press: Oxford.

Cook, M. and G. Wilson (Eds.) (1979). Love & Attraction. Pergamon Press Ltd. Oxford.

Dion, K. K., E. Berscheid, and E. Walster (1972). "What is beautiful is good." Journal of Personality & Social Psychology, 24(3):285-290.

Feingold, A. (1990). "Gender differences in effects of physical attractiveness on romantic attraction." Journal of Personality and Social Psychology, 59(5):981-993.

Feingold, A. (1992). "Good-looking people are not what we think." Psychological Bulletin, 111(2):304-341.

Festinger, L. (1957). "A Theory of Cognitive Dissonance." Standford Univ. Press. California.

Fisher, H. (1992). The Anatomy of Love. Nueva York. W.W.Norton.


Frijda, N. H. (1988). "The laws of emotion." American Psychologist, 43(5):349-358.

Griffitt, W. (1979). "Sexual stimulation and sociosexual behaviors." In M.Cook & G.Wilson (Eds.): Love & Attraction. Oxford: Pergamon Press.

Hadjistavropoulos, T. and M. Genest (1994). "The underestimation of the role of physical attractiveness in dating preferences: Ignorance or taboo?" Canadian Journal of Behavioural Science, 26(2):298-318.

Huston, T. L. and G. Levinger (1978). "Interpersonal attraction and relationships." Annual Review of Psychology, 29:115-156.

Kalick, S. M. and T. E. Hamilton (1986). "The matching hypothesis re-examined." Journal of Personality and Social Psychology, 51:673-682.

Lewis, R. A. (1972). "A developmental framework for the analysis of premarital dyadic formation." Family Process, 11:17-48

Mettee, D. R. and E. Aronson (1974). "Affective reactions to appraisal from others." In T.L.Huston (Ed.): Foundations on Interpersonal Attraction. Academic Press. New York.

Murstein, B. I. (1972). "Physical attractiveness and marital choice." Journal of Personality and Social Psychology, 22:8-12.

Reiss, I. L. (1960). "Toward a Sociology of the heterosexual love relationships." Marriage and Family Living, 22:139-145.

Sangrador, J. L. (1993). "Consideraciones psicosociales sobre el amor romántico [Psychosocial considerations about romantic love]." Psicothema, 5:181-196.

Sangrador, J. L. and C. Yela (2000). "What is beautiful is loved: physical attractiveness in love relationships on a representative sample." Social Behavior and Personality, 28(3):207-218.


Skinner, B. F. (1953). Science and human behavior. New York: Macmillan.

Sternberg, R. J. (1986). "A triangular theory of love." Psychological Review, 93(2):119-135.

Sternberg, R. J. (1988). The triangle of love: Intimacy, Passion and Commitment. New York: Basic Books.

Walster, E., G. W. Walster, and E. Berscheid (1978). Equity: Theory and Research. Allyn & Bacon. Boston. Massachusets.

Wilson, G. and D. Nias (1976). Love's mysteries. Glasgow: W. Collins Sons & Co. Ltd.

Wong, F. Y., D. R. McCreary, C. C. Bowden, and S. M. Jenner (1991). "The matching hypothesis: factors influencing dating preferences." Psychology: A Journal of Human Behavior, 28(3-4):27-31.

Yela, C. (1996). "Componentes básicos del amor: algunas matizaciones al modelo de R.J.Sternberg [Basic components of love: some refinements to the model of R.J.Sternberg]". Revista de Psicología Social, 11(2):185-201.

Yela, C. (1998). "Temporal course of the basic components of love throughout relationships." Psychology in Spain, 2(1):76-8

Yela, C. (2000a). "Predictors of and factors related to loving and sexual satisfaction for men and women." European Review of Applied Psychology, 50(1):235-242.

Yela, C. (2000b). El amor desde la Psicología Social: ni tan libres ni tan racionales [Love from Social Psychology: nor so free neither so rational]. Madrid: Pirámide.

Zajonc, R. B. (1968). "Attitudinal effects of mere exposure." Journal of Personality and Social Psychology, 9:1-29.

Zajonc, R. B. (1987). "Convergence in the physical appearance of spouses." Motivation and Emotion, 11:333-346.


This paper is part of a bigger research project funded by the "Comisión Interministerial de Ciencia y Tecnología" (PB 91-0360), carried out by: Ana Barrón, David Martínez-Íñigo, Pilar de Paúl, Jose Luis Sangrador and Carlos Yela, and supervised by Florencio Jiménez Burillo (in alphabetical order).



Carlos Yela is a Professor in the Social Psychology Department at the Complutense University of Madrid. His main research line deals on the social psychology of love, and on social problems. He wrote his doctoral thesis on the social psychology of love.. He has been working during the last decade, with some other colleagues in his department, on the first empirical study about love ever made in Spain with a representative sample. He has presented many papers about love in different national and international congresses, and has published many articles on the matter in different national and international journals. His book, titled "El amor desde la Psicología Social: ni tan libres ni tan racionales" (Love from Social Psychology: Nor So Free Neither So Rational), has recently been published. E-mail address: (or

José Luis Sangrador is a Professor in the Social Psychology Department at the Complutense University of Madrid. His main research lines deal on Social Psychology and on the social psychology of love. He has been working during the last decade, with some other colleagues in his Department, on the first empirical study about love ever made in Spain with a representative sample. He has presented some papers about love in different congresses, and has published some articles on the matter in national and international journals. E-mail address:


Back to Journal Cover Page