Effects of emotional valence and arousal on recollective and nonrecollective recall.

Abstract

The authors investigated the effects of valence and arousal on memory using a dual-process model that quantifies recollective and nonrecollective components of recall without relying on metacognitive judgments to separate them. The results showed that valenced words increased reconstruction (a component of nonrecollective retrieval) relative to neutral words. In addition, the authors found that positive valence increased recollective retrieval in comparison to negative valence, whereas negative valence increased nonrecollective retrieval relative to positive valence. The latter effect, however, depended on arousal: It was reliable only when arousal was high. The present findings supported the notion that emotional valence is a conceptual gist because it affected nonrecollective retrieval and because subjects' recall protocols were clustered by valence. The results challenge the hypothesis that valence affects only recollection, and they clarify previous inconsistent findings about the effects of emotion on memory accuracy and brain activity.

Decision making and cancer.

“…Memory for valence is retained over long periods, supporting the conclusion that it is represented as gist (if valence were represented in verbatim memory, it would fade quickly). That is, the emotional essence of an experience is retained and, as models of recall show, can be used later to reconstruct details (Gomes, Brainerd, & Stein, 2013; Reyna, 2011). Valence, and also discrete emotions (e.g., fear, anger, happiness, and so forth), act as organizing themes for recalling the past and extrapolating to the future (Rivers et al, 2008).…”

Markovian interpretations of dual retrieval processes

“…In the non-canonical design, testing effects are separated from study effects between the first and second opportunities to retrieve the focal list, and confounded thereafter, as the design becomes canonical. In prior investigations (Brainerd et al, 2009, 2010, 2012; Gomes et al, 2013), however, recall has rarely differed between the first two tests in the non-canonical design, and when it does, which usually only occurs in clinical samples, recall is lower on the second test than on the first test. Contrary to such designs, the delayed repeated recall procedure disentangles testing from study effects.…”

“…Next, we present six parameterizations of the starting vector and transition matrix, all of which have been used in recent research applications (Brainerd et al, 2009, 2010, 2012; Gomes, Brainerd, & Stein, 2013). Each delivers identifiable measures of direct access, reconstruction, and familiarity judgment.…”

“…A canonical design in which subjects receive a minimum of 3 study-test trials (e.g., S 1 T 1 S 2 T 2 S 3 T 3 ) has been shown to identify 6 theoretical parameters of the dual-retrieval model (Brainerd et al, 2009, 2012; Gomes et al, 2013) by either assuming that (a) learning after the first trial only occurs on errors, (b) learning after the first trial only occurs on successes, or (c) learning after the first trial occurs both on errors and successes. All three model versions share the common starting vector $${\mathbf{w}}^{false(1false)}=false[D_{\mathit{1}},falseitalic(\mathrm{italic1}-D_{\mathit{1}}falseitalic)R_{\mathit{1}}falseitalic(\mathrm{italic1}-J_{\mathit{1}}falseitalic),falseitalic(\mathrm{italic1}-D_{\mathit{1}}falseitalic)R_{\mathit{1}}{J}_{\mathit{1}},falseitalic(\mathrm{italic1}-D_{\mathit{1}}falseitalic)falseitalic(\mathrm{italic1}-R_{\mathit{1}}falseitalic)false],$$ and therefore, all three posit that the probability of correctly recalling a target on the first trial (the special case n = 1) is P 1 (Rc) 〈 c,w (1) 〉 = D 1 + (1 – D 1 )R 1 J 1 , as in Equation 1.…”

“…Although there are conceptual distinctions among them (e.g., whether learning occurs on errors or successes), the two retrieval processes and the slave judgment process are the same. Previous studies in which such models were implemented (Brainerd et al, 2009, 2010, 2012; Gomes et al, 2013) have shown that D parameters are more sensitive than R or J parameters to manipulations that should affect recollective retrieval (e.g., list length, associative vs. free recall), while R and J parameters are more sensitive than D parameters to manipulations that ought to affect reconstruction or familiarity (e.g., unrelated vs. categorized lists, neutral vs. negative emotionally-valenced lists).…”