Elsevier

Psychiatry Research

Volume 248, February 2017, Pages 56-63
Psychiatry Research

Psychometric properties of the written version of the autobiographical memory test in a japanese community sample

https://doi.org/10.1016/j.psychres.2016.12.019Get rights and content

Highlights

  • The Autobiographical Memory Test is a widely used measure of memory specificity.

  • Its psychometric properties were examined in a large Japanese sample.

  • We replicated a uni-factorial structure and correlations with depressive symptoms.

  • We found a significant declining trend as a function of age.

  • This reduces the association between memory specificity and depression in elderly.

Abstract

The autobiographical memory test (AMT) is a widely used measure to assess the specificity of autobiographical memories. Reduced Autobiographical Memory Specificity (AMS) or increased overgeneralization of memories is considered as a cognitive hallmark of depression. Therefore, reduced AMS is the subject of much psychopathological research, and is a promising target for psychological interventions. Although considerable evidence has been gathered on the clinical relevance of reduced AMS over the past decades, studies on AMS have been mainly conducted in Western populations, and few have been conducted in Asian populations. This could be because of the unknown psychometric properties of the AMT given cultural and language differences. Therefore, the present study examined the psychometric properties of the AMT in a Japanese community sample (N=1240). Our data replicated that (a) the AMT has a uni-factorial structure; (b) AMS has a small but statistically significant negative correlation with depressive symptoms; (c) AMS shows a significant declining trend as a function of age, which influences the magnitude of the association between AMS and depressive symptoms in older adults. These findings suggest that the AMT has robust psychometric properties across different languages and cultural backgrounds.

Introduction

Reduced specificity in autobiographical memory is recognized as a hallmark of depressive cognition, which has been intensively studied in clinical psychology and related fields (Williams et al., 2007). Since the seminal work of Williams and Broadbent (1986), it has been repeatedly shown that individuals with depression have difficulty retrieving specific events that they experienced in the past (Dalgleish and Werner-Seidler, 2014, Williams et al., 2007). This phenomenon is referred as reduced autobiographical memory specificity or overgeneral autobiographical memory, which is considered as a cognitive vulnerability factor that not only correlates with concurrent levels of depressive symptoms (van Vreeswijk and de Wilde, 2004), but also predicts future levels of depression after controlling for baseline levels (Brittlebank et al., 1993, Peeters et al., 2002, Raes et al., 2006; see Sumner et al., 2010, for a review).

Autobiographical memory specificity (AMS) is typically measured using the Autobiographical Memory Test (AMT; Williams and Broadbent, 1986). In the AMT, respondents are presented with several emotional cue words (e.g., happy, sad), and asked to recall and describe their personal memories of which those cue words remind them. Each response is scored by researchers on the basis of pre-established criteria (Williams and Broadbent, 1986), in which specific memories are operationally defined as memories referring to an event that occurred at a particular place and time, and lasted less than one day. Although the AMT cue words vary across studies, most studies use 10 cue words (including five negative and five positive adjectives) and count the number (or proportion) of “specific” responses as the primary index of AMS (Griffith et al., 2012a; van Vreeswijk and de Wilde, 2004).

The AMT has been administered in a wide range of samples (e.g., clinical and non-clinical populations, children, adolescents, and elderly; Griffith et al., 2012a, Griffith et al., 2012b); however, the evidence on reduced AMS has been gathered in Western societies. Only a handful of studies have investigated reduced AMS in Asian (particularly East Asian) populations (e.g., Dritschel et al., 2011; Kong, et al., 2015; Liu et al., 2016; Matsumoto and Mochizuki, 2016), and in most cases, these rare studies confirm the association between reduced AMS and depressive symptoms in Asian individuals. One possible factor preventing researchers from conducting AMS studies in Asia is that the psychometric properties are still unknown under the potential language and cultural differences between Western and Eastern societies, for example, in terms of self-construal and emotion knowledge (Markus and Kitayama, 1991, Wang, 2011). Because depression is estimated as one of the most serious social problems in Asian countries (e.g., Mathers and Loncar, 2006; Okumura and Higuchi, 2011), it is important to validate the AMT for facilitating cognitive research on depression in Asia, and for documenting and understanding similarities and/or differences in AMS across people with different (cultural) backgrounds.

Therefore, in the present study, we examined the psychometric properties of the AMT in a large Japanese community sample by (a) comparing mean levels of AMS between Japanese vs. Belgian and UK samples (Debeer et al., 2009, Heron et al., 2012); (b) examining the (uni-)factorial structure of AMS across positive and negative cue words (cf. Griffith et al., 2009; Griffith et al., 2012b); (c) testing correlations between AMS and depressive symptoms and related psychopathological variables (i.e., rumination, anxiety, and worry; de Decker, et al., 2003; Raes et al., 2006); and (d) examining the age dependency of AMS as some studies have shown a declining trend for AMS with increasing age (e.g., Brittlebank et al., 1993).

First, it has been shown that Asian (Chinese, Chinese-American, or Taiwanese) people tend to recall autobiographical memories with less specific details than American (e.g., Wang, 2001; Wang et al., 2011) and British individuals (Dritschel et al., 2011); however, another study shows no differences in memory specificity (measured by the sentence completion task; Raes et al., 2007) between Chinese and Australian participants (Chen et al., 2015). The reduced specificity in Asians is argued to be rooted to the unique parental education styles in East Asian cultures, which emphasize social roles rather than personal autonomy. In many East Asian cultures, such as China, Korea, and Japan, people believe that emotion needs to be strictly controlled because emotion is considered as destructive or even dangerous to ongoing relationships and social harmony; however, in Euro-American cultures, expressing emotion is encouraged, as it is an affirmation of individual uniqueness (Wang, 2011). These arguments suggest that Japanese people might generate less specific memories on the AMT than do European and American people, due to the tendency to suppress emotional expressions and emotional memories. This point is particularly important when it comes to the feasibility of the AMT in depression research, because such a general tendency to respond with less specific memories would limit the sensitivity of the AMT to discriminate between individuals with and without vulnerability for depression. Therefore, we compared AMS of a Japanese sample to that of European samples, using the same or a highly similar AMT procedure (Debeer et al., 2009, Heron et al., 2012).

Second, because the AMT often includes positively and negatively valenced cue words, some researchers have scored memory specificity separately for positive and negative items. However, a meta-analytic review suggests that the findings on the “valence effect” of the AMT are not consistent across studies (Van Vreeswijk and de Wilde, 2004); some studies found that patients with depression responded with less specific and more overgeneralized memories to positive than negative cue words (e.g., Park, Goodyer, and Teasdale, 2002), whereas other studies showed reduced specific memories for negative rather than positive cues (e.g., Mackinger et al., 2000). Griffith et al., 2009, Griffith et al., 2012b and Heron et al. (2012) examined the item response functions in the framework of the item response theory, suggesting that the AMT has a uni-factorial structure and does not have positive and negative latent factors. Therefore, they argued that there is no statistical support for calculating separate scores for negative and positive items, and that the valence effect would be more likely to reflect the idiosyncratic differences of the items rather than latent emotional factors. Following these findings, we tested the factor structure of the AMT. We examined whether the AMT has separate “positive” and “negative” factors or a single memory specificity factor.

Third, we examined the associations between AMS and depressive symptoms and related psychopathological variables. Although meta-analyses have shown that reduced AMS is associated with clinical levels of depression (Sumner et al., 2010; Van Vreeswijk and de Wilde, 2004), studies in non-clinical samples often failed to find a significant correlation between AMS and depressive symptoms (Raes et al., 2007). This is because the standard AMT, which explicitly asks respondents to describe a specific memory (i.e., an event that occurred on a specific day), is most sensitive to discriminate between people with low and extremely low levels of AMS (Griffith et al., 2009). Instead, for non-clinical analogue studies, researchers have proposed the minimal- or shortened-instruction variant of the AMT (Debeer et al., 2009, Heron et al., 2012). The minimal instructions do not explicitly instruct respondents that memories should be specific (Debeer et al., 2009); the shortened instructions provide respondents with examples of a specific and categoric memory, and ask them to describe real events, but do not state that the events should have occurred on a single day (Heron et al., 2012). Indeed, analyses of the item response theory suggest that these adapted versions of instructions are most informative for people falling at around the mean or moderate levels of AMS (Griffith et al., 2009, Heron et al., 2012). Furthermore, AMS measured by the minimal-instruction AMT was found to be correlated with depressive symptoms and rumination, in a non-clinical student sample (rs=−.20 and −.28, respectively; Debeer et al., 2009). Because such negative repetitive thinking—often referred to as rumination in depression and worry in anxiety (Ehring and Watkins, 2008, Watkins, 2008)—is a critical factor that acts to exacerbate negative emotions in interaction with reduced AMS (Nolen-Hoeksema et al., 2008; Raes et al., 2006; Williams et al., 2007), we examined in the present study whether AMS, measured by the AMT with the shortened instructions of Heron et al. (2012), is associated with depressive symptoms, depressive rumination, and anxious worry.

Fourth, it also would be important to examine the age-dependency of AMS, as episodic memories are known to be particularly sensitive to the influence of age (e.g., Brickman and Stern, 2009; Nyberg et al., 1996). Previous studies have suggested that older adults tend to respond with less specific memories on the AMT (Phillips and Williams, 1997, Serrano et al., 2007). This is because of age-related cognitive decline such as in working memory, but not because of change (or increase) in depressive symptoms (Brittlebank et al., 1993, Holland et al., 2012, Phillips and Williams, 1997, Ros et al., 2010, Williams et al., 2007). Other gerontological studies have also shown that older adults with and without depression generate specific memories at equally low levels, and that AMS has no significant associations with the severity of either depressive symptoms or depressive rumination (Phillips and Williams, 1997, Ricarte et al., 2015; but see Ricarte et al., 2011). Thus, we hypothesized a significant interaction between age and depressive symptoms predicting AMS, such that depression-related differences in AMS would be smaller for older than for younger people.

Section snippets

Participants and procedures

Participants were community dwellers living in Japan (n=1,240; 620 males and 620 females)1 and were recruited by an online survey company (Macromill Inc., Tokyo, Japan) from their list of potential (more than 1 million) responders. The participants voluntarily participated in the online survey. After providing demographic information about gender, age, and occupation, the

Basic statistics of the AMT

Table 1 shows average proportions of the five response categories of the AMT. Our participants responded with less specific memories (28%) compared with previous studies using the minimal or shortened instructions of the AMT in a Belgian undergraduate sample (53% in Debeer et al., 2009) and UK child sample (37% in Heron et al., 2012). Instead, the proportion of semantic associates (27%) was higher than those found in European data (8% in Debeer et al., 2009; 3% in Heron et al., 2012), whereas

Discussion

The present study examined the psychometric properties of the AMT in a Japanese community sample. First, we found that the mean level of AMS was lower in the current Japanese sample as compared to previous Belgian and UK samples. We believe that this reduced AMS is not because of cultural differences, but can be mostly attributed to the age diversity in our dataset. As additional evidence, we found that (a) the youngest age group had comparable levels of AMS (around 40%) to the previous data

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgments

Keisuke Takano was supported by the Japan Society for the Promotion Science Postdoctoral Fellowships for Research Abroad. Filip Raes was supported by the KU Leuven Research Council Grant PF/10/005.

References (59)

  • A.D. Brittlebank et al.

    Autobiographical memory in depression: state or trait marker?

    Br. J. Psychiatry

    (1993)
  • X.J. Chen et al.

    Chinese and Australians showed difference in mental time travel in emotion and content but not specificity

    Front. Psychol.

    (2015)
  • T. Dalgleish et al.

    Reduced specificity of autobiographical memory and depression: the role of executive control

    J. Exp. Psychol. Gen.

    (2007)
  • E. Debeer et al.

    Associations between components of rumination and autobiographical memory specificity as measured by a minimal instructions autobiographical memory test

    Memory

    (2009)
  • A. de Decker et al.

    Autobiographical memory specificity and trauma in inpatient adolescents

    J. Clin. Child. Adolesc. Psychol.

    (2003)
  • B. Dritschel et al.

    How are depression and autobiographical memory retrieval related to culture?

    J. Abnorm. Psychol.

    (2011)
  • T. Ehring et al.

    Repetitive negative thinking as a transdiagnostic process

    Int. J. Cognit. Ther.

    (2008)
  • F. Gotoh et al.

    Affective valence of two-compound kanji words

    Tsukuba Psychol. Res.

    (2001)
  • J.W. Griffith et al.

    Factor structure of the ruminative response scale: a community-sample study

    Eur. J. Psychol. Assess.

    (2015)
  • J.W. Griffith et al.

    An item response theory/confirmatory factor analysis of the autobiographical memory test

    Memory

    (2009)
  • J.W. Griffith et al.

    The factor structure of the autobiographical memory test in recent trauma survivors

    Psychol. Assess.

    (2012)
  • J. Heron et al.

    40,000 memories in young teenagers: psychometric properties of the Autobiographical Memory Test in a UK cohort study

    Memory

    (2012)
  • C.A. Holland et al.

    Executive function and emotional focus in autobiographical memory specificity in older adults

    Memory

    (2012)
  • J.A. Krosnick

    Response strategies for coping with the cognitive demands of attitude measures in surveys

    Appl. Cognitive Psych.

    (1991)
  • Y. Kyung et al.

    Specificity and detail in autobiographical memory: Same or different constructs?

    Memory

    (2016)
  • M. Lamar et al.

    Longitudinal changes in verbal memory in older adults: distinguishing the effects of age from repeat testing

    Neurology

    (2003)
  • Liu, Y., Zhang, F., Wang, Z., Cao, L., Wang, J., Na, A., et al., 2016. Overgeneral autobiographical memory at baseline...
  • H.F. Mackinger et al.

    Autobiographical memories in women remitted from major depression

    J. Abnorm. Psychol.

    (2000)
  • H.R. Markus et al.

    Culture and the self: implications for cognition, emotion, and motivation

    Psychol. Rev.

    (1991)
  • Cited by (25)

    • Different retrieval mechanisms of overgeneral autobiographical memory for positive and negative cues in remitted major depressive disorder

      2023, Journal of Behavior Therapy and Experimental Psychiatry
      Citation Excerpt :

      Categoric memories tend to be proportionally higher on the AMT in depressed groups compared to non-depressed, and previous studies have shown that is evidence for both positive and negative cue words (Liu et al., 2013; Ono et al., 2016; van Vreeswjik & de Wilde, 2004). Therefore, it has been assumed that overgeneral/categoric responses occur regardless of the emotional valence of cue word (Griffith et al., 2009, 2012; Takano, Mori, Nishiguchi, Moriya, & Raes, 2017). However, by distinguishing between generative and direct retrieval, Matsumoto, Takahashi, and Kawaguchi (2020) showed that the mechanisms of OGM in depressed individuals are in fact different for positive and negative cues.

    • Upward action promotes selective attention to negative words

      2021, Heliyon
      Citation Excerpt :

      The CES-D is a self-report questionnaire that includes 20 items rated on a 4-point Likert scale, measuring the feelings of the participants during the past week (score range: 0 to 60). The CES-D score (M = 12.93, SD = 8.41) was comparable to the previous study on healthy Japanese community dwellers (M = 14.11, SD = 10.13; N = 1181; Takano et al., 2017) as the difference between these samples was small (Cohen's d = 0.13). Each participant provided written informed consent before the experiment.

    • The longitudinal association between individual differences in recall of positive specific autobiographical memories and daily cortisol

      2021, Biological Psychology
      Citation Excerpt :

      To our knowledge, there is no precedent for such a ratio score within the literature other than Askelund et al. (2019). Psychometric analyses suggest that individual differences in autobiographical memory specificity are best operationalised as either the number of specific memories retrieved or the proportion of specific memories retrieved relative to the number of cues given (Griffith, Kleim, et al., 2012; Griffith, Sumner, et al., 2012; Heron et al., 2012; Takano et al., 2017). Future studies in this area should continue to operationalise specificity in line with this established precedent so that studies regarding the association between cortisol dysregulation and specificity can align with broader evidence base regarding autobiographical memory specificity.

    • Efficacy of online Memory Specificity Training in adults with a history of depression, using a multiple baseline across participants design

      2019, Internet Interventions
      Citation Excerpt :

      Future studies examining c-MeST should include a control group to account for this. An additional limitation of the current study is that we used a verbal AMT with feedback, in contrast with studies in which test-retest reliability of the AMT was examined using versions of the AMT without feedback (i.e. Takano et al., 2017a). This makes it difficult to compare pre-intervention specificity scores with specificity scores of other studies.

    • Lost in distractors: Reduced Autobiographical Memory Specificity and dispersed activation spreading over distractors in working memory

      2017, Behaviour Research and Therapy
      Citation Excerpt :

      Moreover, we only used positive stimuli as targets to eliminate potential contamination by valence effects. Although we cannot completely exclude the possibility that the attention and/or working memory function varies across negative and positive materials, this seems quite unlikely given that AMS is not influenced by valence differences of the cue words (Griffith, Kleim, Sumner, & Ehlers, 2012; Griffith et al., 2009; Heron et al., 2012; Takano et al., 2017; for a review, see Griffith, Sumner, Raes, Barnhofer, Debeer, & Hermans, 2012). Taken together, our results indeed show that AMS is associated with inhibition of activation for cue-unassociated distracters, but it is still unclear whether AMS relies on selective activation for specific representations of individual event memories, and whether the inhibition is influenced by valence effect.

    View all citing articles on Scopus
    View full text