Elsevier

Acta Psychologica

Volume 98, Issues 2–3, March 1998, Pages 291-310
Acta Psychologica

The influence of attention at encoding on direct and indirect remembering

https://doi.org/10.1016/S0001-6918(97)00047-4Get rights and content

Abstract

The relation between attention at encoding and direct (i.e., recognition) versus indirect (i.e., rapid reading) remembering was investigated. In Experiments 1 and 2, color of print indicated whether to read an individual word aloud or to ignore it. This attentional manipulation reduced direct but not indirect remembering for the ignored words relative to the attended words. Apparently direct remembering is extremely dependent on attention at encoding. In Experiment 3, however, presenting two words simultaneously at study, with color now signifying which word to read and which to ignore, eliminated this dissociative effect of attention. Ignored words were not remembered on either test, although attended words were remembered well on both. Mere exposure is not sufficient to produce indirect remembering: Stimuli must be attended. Ignoring one stimulus in favor of processing another stimulus that is simultaneously presented and equally salient may prevent even the minimal attentional requirements of indirect remembering from being met, let alone the more stringent requirements of direct remembering.
PsycINFO classification: 2343

Introduction

To perform well on traditional tests of memory, subjects must consciously recollect previously encoded events (Johnson and Hasher, 1987; Richardson-Klavehn and Bjork, 1988; Roediger and McDermott, 1993). Free recall and recognition tests are the hallmark examples of such tests, which have long been the primary tools for the study of memory. These tests may be referred to as direct tests, in that subjects are encouraged to make explicit contact with one or more previous encoding episodes. In the past 15–20 years, however, memory researchers have become increasingly intrigued by another class of memory tests: indirect tests.

On indirect tests of memory, subjects are not asked to refer back to a prior episode, despite its relevance to the task at hand. Often, though not necessarily, the relevance of the prior episode or study phase is even concealed from subjects. In these tasks, evidence of retention of the encoding episode is usually revealed by facilitated (i.e., faster or more accurate) performance for items that were previously presented compared to equivalent items that were not previously presented. This facilitated performance is usually referred to as priming or repetition priming. Generally, the instruction to refer or not to refer to previously presented material has been used as an operational distinction between direct and indirect tests (Roediger and McDermott, 1993).

There are numerous examples of indirect tests of memory. The word fragment completion task (Tulving et al., 1982) involves presenting words with letters omitted and requiring subjects to fill in the missing letters. When a solution to a fragmented word was previously presented, subjects more often offer correct completions. In the masked word identification task (also referred to as the perceptual identification task; Jacoby and Dallas, 1981), words – some previously studied and some new – are presented to subjects very briefly and then masked. Subjects are usually able to identify more studied words than unstudied words despite no instructions to use memory or to refer to the study list (e.g., Masson and MacLeod, 1992). A further example of an indirect test is the rapid reading procedure (also called pronunciation or naming; e.g., Scarborough et al., 1979) where latencies to read words that were previously presented are normally shorter than are latencies to read new, unstudied words (e.g., MacLeod, 1996). All of these tests have been used as indirect measures of memory, along with an ever-growing battery of others.

Attentional manipulations at encoding have been found to affect direct but not indirect tests of memory (Parkin et al., 1990; Szymanski and MacLeod, 1996). Research shows that dividing or diverting attention from items at study impairs performance on subsequent direct tests but does not similarly impair performance on indirect tests. This dissociation will be the focus of the experiments that we will present. Before describing our experiments, we will review previous investigations of how attention at study affects performance on direct versus indirect tests of memory.

Research has begun to emphasize the interplay between attention and memory as measured by different types of tests. Early on, Eich (1984)used a dichotic listening procedure, having participants shadow prose to the attended ear while hearing word pairs consisting of a homophonic noun preceded by an adjective that biased its interpretation (e.g., taxi-FARE) in the unattended ear. Although on a subsequent recognition memory test participants did not recognize the words from the unattended ear, they demonstrated the appropriate bias in an oral spelling test (e.g., spelling “fare” rather than “fair”). This finding reveals preserved priming on an indirect test for items that were not attended, despite no evidence of memory for these items on a direct test.

Parkin et al. (1990)examined the effect of divided attention at study by asking participants in a full attention condition to complete a sentence verification task only and participants in a divided attention condition to perform sentence verification and tone monitoring tasks simultaneously. The sentence verification task involved reading sentences and judging whether they made sense; the tone monitoring task consisted of listening to a series of tones occurring every 3–7 s and indicating for every tone whether it was high, medium, or low frequency. After a 24-h delay, participants performed a recognition test and a word fragment completion test. As expected, Parkin et al. found recognition to be much poorer under divided attention conditions than under full attention conditions. On the word fragment completion test, however, divided attention at encoding had no ensuing effect: both groups showed equivalent priming for studied words. These results suggest that implicit remembering occurs despite constraints on attentional resources at study. Divided attention at study is not so kind, however, to performance on direct tests, which suffers substantial impairment. This conclusion was strengthened by the corresponding findings of Parkin and Russo (1990)using picture completion as both the direct and the indirect measure of memory.

Szymanski and MacLeod (1996)also examined the impact of selective attention at study on subsequent direct and indirect memory tests. During encoding, participants read some words aloud and named the print colors of others. In the color-naming study phase, the word was ignored whereas the color was attended. At test, participants either performed a direct recognition test or an indirect lexical decision test. In both tests, words from the study phase appeared among new words. Szymanski and MacLeod found that words that were read at study were better recognized than words that were color-named at study. The repetition priming in the lexical decision task, however, was equivalent for words that were read and words that were color-named at study. Szymanski and MacLeod concluded that successful implicit remembering occurs once a low attentional threshold is met, and that further attention confers little additional benefit. Their study provides yet another illustration of dissociated performance on direct and indirect tests of memory.

Although many studies have found preserved priming on an indirect test despite impaired performance on a direct test following reduced attention at study, recent evidence has emerged indicating that the effect of attention at encoding on direct and indirect remembering is not as straightforward as it first seemed.2 Noteworthy among these, Wood and Cowan (1995)investigated both types of tests in a series of selective listening experiments. Participants were asked to ignore passages presented to one ear while attending to and shadowing passages presented to the other ear. A segment of backward speech was embedded in the passage presented to the unattended ear, and participants were later asked to indicate whether they could recall anything odd happening in the unattended ear. Although some did remember hearing a switch to backward speech, their recollection occurred at a cost to on-line shadowing. For participants who detected the shifts, errors in shadowing occurred about 15 s following the shift from forward to backward speech in the unattended ear. Evidently, remembering did not occur in the absence of attention at study. In addition, both direct and indirect tests of memory revealed little to no memory for the content of the unattended message in which the backward speech had been embedded. These findings suggest that success on both direct and indirect tests requires attention at encoding.

In his Experiment 10, Phaf (1994)also investigated the effect of attention at encoding on direct and indirect remembering. Participants were presented two words simultaneously on each trial, one to the left and one to the right of fixation. After a 200-ms delay, an arrow appeared pointing to one word, instructing participants to read it aloud. Participants were to attend only to the cued word, and were urged to ignore the other word. Three tests were later administered. One group did a word-stem completion test, with unique three-letter cues corresponding to studied or unstudied words. Another group did a masked word identification test, where each test word briefly appeared inside a mask. Finally, a third group did a free recall test. The free recall test showed essentially no memory for the ignored words, consistent with the need for prior attention on direct tests. More surprisingly, however, both indirect tests also demonstrated impairment for words that were ignored at study. Even though there was still reliable priming for the ignored words, participants showed more priming for attended words. In contrast to the previously described results, these findings suggest that attention may play a role in both direct and indirect remembering.

The most common finding is that there is virtually no effect of decreased attention to items at study on later indirect tests despite seriously detrimental effects on later direct tests (Eich, 1984; Szymanski and MacLeod, 1996). Although this pattern is not uncontested, it suggests that success on indirect tests does not depend on attention at encoding despite the fact that success on direct tests depends greatly on it. The following contention expresses this view about indirect remembering: “...the memory processes underlying priming effects are not open to conscious inspection and should not be disrupted by experimental manipulations that reduce the subjects' degree of conscious involvement with the initial learning task” (Parkin et al., 1990, p. 510). Others also argue that indirect remembering is independent of attention and that automatic sensory registration or data-driven processing is sufficient to facilitate reprocessing of studied items on indirect tests (Hayman and Tulving, 1989; Roediger, 1990).

The objective of the present study was to explore further this interplay between attention and memory. The aim was to increase clarity concerning the contribution of attention at study to subsequent remembering as measured by these two classes of test, direct and indirect. Do all reductions in attention at encoding, no matter how slight, produce declines in performance on direct tests? Do all reductions in attention at encoding, no matter how extreme, fail to affect indirect remembering? Experiments 1 and 2 investigated whether direct remembering would be affected by a subtle attentional manipulation. The goal was to examine the sensitivity of direct remembering to even slight reductions of attention during encoding. In Experiment 3, the primary objective was to create a dramatic attentional manipulation at study to determine whether indirect remembering would be reduced if the manipulation of attention was sufficiently extreme.

Section snippets

Experiment 1

The aim of Experiment 1 was to investigate the effect of a subtle attentional manipulation at encoding on direct and indirect remembering. In most previous studies, attention to items has been limited either by competing stimuli (i.e., selective attention situations) or by competing responses (i.e., divided attention situations). In these cases, the attentional limitations on processing of tested items have been quite severe and usually have resulted in seriously impaired performance on direct

Experiment 2

The objective of Experiment 2 was to re-examine the effect of an attentional manipulation at encoding on performance on direct and indirect tests of memory. The result on the direct test of Experiment 1 could simply have been due to impeded encoding of white words because of the irrelevant `pass' response that was made during study of these items. In the present experiment, participants were instructed to read red words aloud but not to read white words aloud, thereby removing this impediment.

Experiment 3

The aim of Experiment 3 was again to limit attention to some words at study and to observe how this affected performance on both direct and indirect tests, with the specific goal of altering performance on the indirect test. The intent was to investigate whether indirect remembering could be affected at all by reductions in attention at study or whether encoding for this type of remembering truly occurred automatically without attention to items at encoding (Hayman and Tulving, 1989; Parkin et

General discussion

Not surprisingly, the role that attention plays in memory is complex. The emphasis in the present study was on the influence of attention at the time of encoding on subsequent memory as measured by two predominant categories of memory tests, direct and indirect. Previous work (Eich, 1984; Parkin et al., 1990; Parkin and Russo, 1990; Szymanski and MacLeod, 1996) led to the conclusion that whereas variation in attention at encoding certainly affected performance on direct memory tests, it had

Acknowledgements

This research was supported by Natural Sciences and Engineering Research Council of Canada grant A7459. The three experiments form part of the senior author's Master's thesis. We appreciate the helpful comments of Gordon Logan and Gezinus Wolters on the initial version of this article.

References (35)

  • C.M MacLeod

    How priming affects two speeded implicit tests of remembering: Naming colors versus reading words

    Consciousness and Cognition

    (1996)
  • K.F Szymanski et al.

    Manipulation of attention at study affects an explicit but not an implicit test of memory

    Consciousness and Cognition

    (1996)
  • C.B Boronat et al.

    The role of attention in automatization: Does attention operate at encoding, or retrieval, or both?

    Memory and Cognition

    (1997)
  • A.S Brown et al.

    A reevaluation of semantic versus nonsemantic processing of implicit memory

    Memory and Cognition

    (1994)
  • B.H Challis et al.

    Level of processing affects priming in word fragment completion

    Journal of Experimental Psychology: Learning, Memory, and Cognition

    (1992)
  • J.A Debner et al.

    Unconscious perception: Attention, awareness, and control

    Journal of Experimental Psychology: Learning, Memory, and Cognition

    (1994)
  • E Eich

    Memory for unattended events: Remembering with and without awareness

    Memory and Cognition

    (1984)
  • E Fox

    Negative priming from ignored distractors in visual selection: A review

    Psychonomic Bulletin and Review

    (1995)
  • R Graves et al.

    Millisecond interval timer and auditory reaction time programs for the IBM PC

    Behavior Research Methods, Instruments, and Computers

    (1987)
  • R Graves et al.

    More on millisecond timing and tachistoscope applications for the IBM PC

    Behavior Research Methods, Instruments, and Computers

    (1988)
  • C.A.G Hayman et al.

    Is priming in fragment completion based on a “traceless” memory system?

    Journal of Experimental of Psychology: Learning, Memory, and Cognition

    (1989)
  • L.L Jacoby et al.

    On the relationship between autobiographical memory and perceptual learning

    Journal of Experimental Psychology: General

    (1981)
  • M.K Johnson et al.

    Human learning and memory

    Annual Review of Psychology

    (1987)
  • A.J Lambert et al.

    Selective attention, visual laterality, awareness, and perceiving the meaning of parafoveally presented words

    Quarterly Journal of Experimental Psychology: Human Experimental Psychology

    (1988)
  • G.D Logan et al.

    What is learned during automatization? The role of attention in constructing an instance

    Journal of Experimental Psychology: Learning, Memory, and Cognition

    (1994)
  • G.D Logan et al.

    Attention in the acquisition and expression of automaticity

    Journal of Experimental Psychology: Learning, Memory, and Cognition

    (1996)
  • MacDonald, P.A., 1997. The Influence of Attention at Encoding on Direct and Indirect Remembering. Unpublished Master's...
  • Cited by (93)

    • Neural correlates of the production effect: An fMRI study

      2021, Brain and Cognition
      Citation Excerpt :

      Stimuli were presented using a custom script built in PsychoPy2 1.84.2 (Peirce, 2009). The words were 120 nouns from MacDonald and MacLeod (1998), 5 letters to 10 letters in length, with frequencies greater than 30 per million (Thorndike & Lorge, 1944). For each subject, the script randomly assigned each word to one of four lists (30 words each), corresponding to the four experimental conditions (read aloud, read silently, sensorimotor control, and foils).

    • The impact of different background noises on the Production Effect

      2018, Acta Psychologica
      Citation Excerpt :

      Accordingly, vocal production provides a boost to memory by selectively focusing attention on aloud over silent words (Ozubko, Hourihan, & MacLeod, 2012). MacDonald and MacLeod (1998) stated that “words that received more attention at encoding were better recognized than words that received less attention. The more thorough processing of words that were read aloud at study led to better direct remembering.” (

    • Actions do not clearly impact auditory memory

      2023, Frontiers in Human Neuroscience
    View all citing articles on Scopus
    View full text