In Hazan & Shaver there is no way of knowing if the same individuals who report secure adult relationships were also securely attached as children.
The results of Hazan & Shaver support the 'continuityhypothesis' of attachment styles.
Hazan & Shaver had 620 responses which is a good sample size to generalise from.
Retrospective data about childhood attachments is unreliable.
Asking people to recall information from the past is unreliable and subjective.
The love quiz may have had sociallydesirable answers.
The sample was self-selected so may be biased.
More females than males responded to the love quiz (205 males, 415 females) making it genderbiased.
The study provided a bridge between infant attachment theory and theories of romantic love, which is very useful.
The findings have been replicated (reliable) and researchers have linked adult attachment to existing theories of love.
Fearon & Roisman (2017) also concluded that early attachment predicts later attachment.
Becker-Stoll et al (2008) followed 43 children from one year of age, and at 16 found no evidence of continuity.
Validity of longitudinal studies can be decreased if internal measures are not standardised.
Individualdifferences will always play a part in any study into attachment and future relationships.
Kirkpatrick & Davis (1994) studied dating couples and found a positive correlation between early attachment and satisfaction in their current relationship.
Brennan & Shaver (1995) found children with insecure avoidant relationships were more likely to engage in casual relations.
More research in this area is needed. Steele (1998) only found a small link between childhood and adult attachments.
McCarthy (1999) studied 40 women who had been assessed for attachment style while in infancy and found that those who were secure in infancy were more likely to be in secure adult relationships.