Memory Effects & The Working Memory Model

Want to improve your memory? The secret is to understand your working memory. In this post, we’ll review some factors that can reduce memory and then explain these using Baddeley and Hitch’s working memory model.

The Articulatory Suppression Effect

The articulatory suppression effect happens when memory decreases because researchers block (suppress) someone’s ability to rehearse (articulate) information they’re trying to remember.  Studies using this technique will give participants a random list of words or letters to remember, but instruct them to repeat something else out loud (e.g. saying “the” repeatedly). Participants forget what they were trying to remember.

Key Study: Alloway et al. (2010) asked participants to remember sequences of numbers. There was a control condition, two verbal suppression conditions (quietly repeating the letter “b” or repeating ‘‘a-b-c-d-e-f-g’’ at a rate of two letters per second) and two visuo-spatial suppression conditions (tapping on a single point on a table with their hand twice per second or tapping four dots on the table in a clockwise sequence). The results showed the verbal suppression reduced memory and the more complicated sequence had the biggest effect. The same reduction in memory wasn’t observed in the visuospatial condition. However, in a second experiment the procedures were the same but participants had to remember visual items (sequences of grey blocks). The results were reversed – the visuospatial suppression reduced memory but not articulatory suppression.

This is an example of a dual task study where participants are asked to do two different tasks at once. These support the idea that working memory has two distinct parts that process different types of information.

The Phonological Similarity Effect

The phonological similarity effect is when participants find it harder to remember the exact order of items that sound similar (e.g. B, C, T, E, G) to those that sound quite different (X, Z, A, R, W). This was first discovered by Conrad in 1964, which was one of the main studies that inspired the working memory model. The effect has been replicated numerous times and is the primary support for the working memory model.

Articulatory Suppression and the Phonological Similarity Effect

But what happens when you test phonological similarity and use articulatory suppression? If you present the items to be remembered visually, then magically the phonological similarity effect disappears!

Tip: These effects aren’t being explained yet to encourage you to explain them for yourself using the working memory model.

The Word Length Effect

This one’s pretty easy – it’s harder to remember long words than short ones. More specifically, words that take longer to say are remembered more poorly than those that take less time.

Key Study: Baddeley et al. (1975): In one study, participants were asked to remember words of varying lengths: one group of eight monosyllable words (sum, hate, harm, wit, bond, yield, worst, and twice) and one group of eight five-syllable words (association, opportunity, representative, organization, considerable, immediately, university, and individual). The shorter words were remembered better.

From the original study by Baddeley et al. 1975.

Irrelevant effects

The irrelevant sound effect happens when distracting sounds reduce memory. It’s not just any sound, though. If a sound is constant it’s not as distracting as a random sound. Have you ever been studying while listening to music and then suddenly an advertisement breaks your attention? It’s the same thing. The random irrelevant sounds distract our attention more-so than constant irrelevant sounds, thus reducing memory.

Similarly, the irrelevant picture effect happens when distracting images are shown while people are trying to remember information using visual mnemonics. 

Key Study: Quinn and McConnell (1996): This study tested the memory of 36 college students. Participants were given a list of words to remember using either rote rehearsal (repeating in your mind) or a visual mnemonic* (creating a mental image of the word in their minds). The words to remember were shown visually but at the same time distracting, irrelevant visual information was also shown. This was drawings that changed every 4 seconds or images of white noise. The results showed that the distracting pictures reduced memory. Moreover, they had a greater effect on using visual mnemonics than rote rehearsal. Finally, the drawings had a greater effect than the white noise. This demonstrates the irrelevant pictures effect on memory.

Explaining the Effects

The Working Memory Model

To understand these effects and studies one must first understand the working memory model (WMM). To recap, the WMM proposes that our short-term memory (the information we’re thinking about right now) has four major components: the central executive (controls our attention), the phonological loop (for sounds), the visuospatial sketchpad (for images), and the episodic buffer (holds information temporarily until we need it).

Since the original WMM in 1974, most of the research has been on the phonological loop. There are two parts to the loop – the phonological store (the inner ear; stores information) and the articulatory loop (the inner voice; the voice in our head when we’re remembering something). According to Baddeley, the phonological store holds information for about 2 seconds before it’s lost. It’s the job of the articulatory loop* to keep refreshing the contents of the store through rehearsal. The effects relating to sounds can be explained by looking at these claims regarding the phonological loop.

According to Buchsbaum (2016), “the phonological loop model of verbal working memory has stood the test of time, largely because it explains many of the behavioral phenomena [e.g. memory effects] associated with verbal memory performance in a simple and intuitive way.”

Explanations

Articulatory Suppression: Blocking rehearsal through articulatory suppression prevents the articulatory loop from being able to keep the information in the phonological store. This is why the effect occurs even with immediate free recall (participants can’t remember words immediately after hearing them). This supports the idea that the articulatory loop is needed to keep information in the phonological store. This is different to the idea of the multi-store model which stated information in the short-term store lasts about 15 to 30 seconds, regardless if it’s rehearsed or not.

Alloway et al.’s study demonstrates the articulatory suppression effect. It also supports the claim that working memory has two components: one for visual information and one for verbal information. If working memory was processed in one unitary store, different types of distracting tasks (verbal or visual) would have the same effects. Alloway et al. claim their results “…can be explained by the multicomponent working memory model: Verbal resources are supported by the phonological loop; spatial resources are managed by the visuospatial sketchpad.”

Phonological Similarity Effect: The similar sounds interfere with one another while they’re being rehearsed in the articulatory loop. The memories become jumbled up and difficult to recall. If no such loop existed, there’d be no differences in our memory for similar or different sounding words.

Suppression Removes the Similarity Effect: If the reason we can’t remember similar sounding things is because they’re jumbled up in our articulatory loop, then if we prevent people from using this loop then the effect will disappear. That’s exactly what the studies have shown.

Word Length Effect: This happens because longer words simply take longer to rehearse in our phonological loop so they get fewer rehearsals. This leads to greater decay. Studies on this use immediate free recall, so it’s not testing the effects of transfer to long-term memory but whether the information is still in the phonological store, i.e. short-term memory.

Irrelevant Sound Effect: This effect is studied when the items to be remembered are presented visually. It is the job of the articulatory loop to transfer these to sounds (grapheme to phoneme conversion; in other words you see the shape of the letter “C’ and then your articulatory loop makes the sound in your head, like “sea” and it’s the sound you rehearse). The irrelevant sounds disrupt this process – the sounds go straight to the articulatory loop and interferes with the rehearsal. Thus, the effect disappears under articulatory suppression – because the participants can’t use their articulatory loop to rehearse the information (they’re too busy repeating the word “the”), the irrelevant sound effect disappears. The same goes for the phonological similarity effect – the suppression of the articulatory loop means the similar sounding items aren’t getting jumbled up in the store.

Another explanation is that the irrelevant sounds distract our attention, so the central executive is disrupted from giving orders to the phonological loop to process the information. This is why sounds that change in pitch and frequency are more distracting than constant sounds.

Irrelevant Picture Effect: The results of Quinn and McConnell (1996) showed that a distracting, irrelevant image (the drawings) reduced memory when using mnemonic (visual) memory techniques, but not verbal, rote rehearsal techniques. This is because the rote rehearsal is handled in the phonological loop, whereas the mnemonics are handled in the visuospatial sketchpad.

Both the irrelevant effects reduce working memory performance on tasks because the irrelevant sounds and pictures distracts our attention of the central executive. Since it’s the role of the central executive to control what happens in the slave systems of working memory, if it’s distracted then performance will decrease.

*A visual mnemonic is a memory technique which involves associating something with an image that will help you remember it. E.g. a Bactrian Camel’s back is shaped like a B whereas a Dromedary’s back is shaped like a D. 

*The articulatory loop has many names, including the articulatory process and articulatory control process. 

References

Buchsbaum, Bradley. (2016). Working Memory and Language. 10.1016/B978-0-12-407794-2.00069-9.

Alloway, T. P., Kerr, I., & Langheinrich, T. (2010). The effect of articulatory suppression and manual tapping on serial recall. European Journal of Cognitive Psychology, 22(2), 297–305. https://doi.org/10.1080/09541440902793731

Baddeley, A. D., Thomson, N., & Buchanan, M. (1975). Word length and the structure of short-term memory. Journal of Verbal Learning and Verbal Behavior, 14(6), 575–589. doi:10.1016/s0022-5371(75)80045-4 (Original)

Quinn, J. G., & McConnell, J. (1996). Irrelevant pictures in visual working memory. The Quarterly journal of experimental psychology. A, Human experimental psychology, 49(1), 200–215. https://doi.org/10.1080/713755613