Short-Term Memory as a Function of Input Modality

Abstract

The two experiments reported are concerned with short-term memory for digit lists simultaneously presented both auditorily and visually. Results showed (1) that interpolated written and verbal recall differentially affect retention depending on whether the to-be-recalled list was presented auditorily or visually. (2) That input modality appears to be far more important for recall than was directing subjects' attention to a list during input, when that list might or might not have been subsequently required for recall. The results suggest that short-term storage is modality specific. In this case, Broadbent's P and S mechanisms do not adequately describe what happens during simultaneous visual and auditory presentation. Nor would Sperling's suggestion of a final auditory store appear to be supported.

Using both an auditory and a visual short-term store to increase digit span

“…The question is whether speaking digits in the auditory store creates less interference with the contents of visual shortterm memory (or whatever nonauditory memory was being used to store the visually presented digits). Speaking might interfere selectively with the contents of auditory short-term memory, as suggested by Margrain's (1967) results. FRICK Second, if dual storage was achieved, the two shortterm stores were storing different quantities of digits.…”

“…This was expected as an example of modality-specific interference, in which speaking is an auditory act that wouldinterfereselectively with auditory short-termmemory. Margrain (1967) found that the recall of auditorily presented digits was more impaired by spoken than by writtenrecallof visually presented digits, and the recall of visually presented digits was more impaired by written than by spoken recall of auditorily presented digits. In Experiment 1, responses werespoken, so an inverted responsebeginning with the auditoryportion was expected to avoid interference.…”

Near-independent capacities and highly constrained output orders in the simultaneous free recall of auditory-verbal and visuo-spatial stimuli.

“…The IFR task instructions gave participants the freedom to output the two types of stimuli in very different output orders if they so wished (e.g., recall the words in forward order starting with the first word, but initiate recall of the visuo-spatial lists with one of the last few dots), but this did not frequently occur. We had also anticipated that participants might output from one modality prior to recall from the other (cf., Broadbent, 1954;Dornbush, 1968;Margrain, 1967), but there was little or no evidence of within-modality clustering. Somewhat to our surprise, we found that the output orders in the recall of auditory-verbal and visuo-spatial stimuli were highly constrained.…”

“…In contrast, previous research has shown that participants might rather prefer to cluster their outputs by stimulus modality. For example, classic early work examining the dichotic presentation for free recall of two separate 3-digit sequences presented to each ear (Broadbent, 1954; for related bisensory serial recall data, see Dornbush, 1968;Margrain, 1967) showed a preferred tendency to recall the items presented to one ear before outputting items presented to the other ear, particularly at faster presentation rates. In addition, when auditory and visual words are presented mixed in a single list, there tends to be an auditory advantage, and clustering by modality during recall (Murdock & Walker, 1969).…”

Four-channel effects in short-term memory

“…However, their proper interpretation is not yet a settled issue. One possibility (Margrain, 1967;Murdock & Walker, 1969) is that there are separate auditory and visual short-term stores and that these modality effects tell us something about the characteristics of these stores. Evidence consonant with this point of view has been reported recently by Fisher & Karsh (1971), Kroll, Parks, Park inson , Bieber, & Johnson (1970), and Parkinson, Parks, & Kroll (1971).…”