BLUCK ET AL.
that this was an initial attempt, it is gratifying that the results of the factor analysis showed some success in mapping the items that we generated from the literature, onto the broad theoretical constructs. Recall that these items were not designed to cohere nicely and tap three functions of AM. Instead, we used the approach of gathering ideas for items wherever we could find mention of proposed func- tions in the theoretical literature. Given that the items were derived in this manner, the variance explained is reasonable. In future scale development, subscales can be “built on” to improve their measurement qualities.
Another issue with the current study is that we set out to see whether the three broad functions cited in the literature could be ver- ified empirically. However, this does not preclude the possibility that there are functions of AM beyond the three that drove this inves- tigation (Bluck, 2003). For example, recent studies of how emotion regulation occurs during memory sharing suggest that emotion reg- ulation may be a separate function of AM (Pasupathi, 2003). It also is unclear whether self–continuity is the only important self function. A series of studies have shown that self–enhancement (e.g., Wilson & Ross, 2003) may also be a function of AM (though this still falls within the broad category of self functions). Thus, we do not claim to have an exhaustive list of functions represented here. As the litera- ture expands it may be that other functions of AM beside the basic directive, self, and social come to light.
Research on the functions of AM can take many forms (see Bluck & Alea, 2002). As mentioned above, using methods that do not rely on self–reports is crucial (for an example, see Alea & Bluck, 2003). Within the realm of self–report, however, one direction we are pur- suing is to use the findings from the current data to revise and update the TALE and to administer it to a larger, and age diverse, sample. Based on the EFA obtained in the current study, and with the larger sample size, confirmatory factor analyses will be conducted to verify the existence of the factors reported here. In addition, collecting the TALE from both older and younger adults will allow us to examine the age invariance of the factor structure across these two groups. One aim in further developing the TALE in this manner is to produce