Glanzer and Cunitz (1966)

Evidence supports multi-store model (MSM)

Multi-Store Model (MSM)

Human memory are divided into 3 parts: has different duration and capacity

1. Sensory memory

   • Requires attention to transfer the info to STM

   • Function: detact and transfer (not processing the info)

   • Sub-components (visual, auditory, …)

   • Duration: several seconds depends on the component

   • Capacity: infinite

2. Short-term memory store

   • Capacity: 7±2 chunks of information

   • Duration: 30 seconds

   • Requires rehearsal to transfer to LTM

3. Long-term memory store

   • Capacity: infinite

   • Duration: infinite

Multi-Store Model’s Strength

* Lots of support evidences
* Most influential memory model
* Able to explain multiple observed phenomenon

Multi-Store Model’s Limitation

* Too focus on the structure compared to process
* Only one way direction for the information to flow (sensory → LTM) which is not true in many cases
* Oversimplify:
* More way for information to transfer
* STM and LTM can be divided further

Serial Postion Effect

Participants can recall the first and last item on the list better than the one in the middle.

* The items in the beginning already go to the longterm memory as it has been repeated too many time
* The items in the end are still in the short-term memory

Aim

To investigate the serial position effect with and without interference from a filler activity.

Participants

46 army enlisted men

Method

* Repeated Measured Design
* Experiment

Procedure

Participants were read a series of 15-word lists and needed to recall the list of words in any order. They were to do this for 3 conditions (5 lists per condition → total of 15 lists is used) :

1. Recall immediately after

2. Recall after a 10 second filler activity

3. Recall after a 30 second filler activity

A filler activity could include counting backwards from a number; the point was to prevent rehearsal. Each participant went through all the conditions, completing 5 lists per condition (15 total). Conditions were randomized.

Result

Condition 1: show serial position effect

* Remembers words at the start → primary effect
* Remembers words at the end → recency effect

Condition 2:

* Still remembers words at the start → primary effect preserved
* No longer remembers words at the end → no recency effect

Conclusion

* Since people intend to memorize the words, they repeat them in their head → words at the start were rehearsed multiple time → longterm memory → unaffect by the filler task


* However the 30 seconds filler task remove word at the end from their short-term memory → unable to recall in the 2nd condition

\
Since only recency effect disappears → show that short term memory and long term memory are separated → supporting multi-store model

Strengths

* All words are one syllable → easy to process by most participant → low participant variability → high construct validity
* Large sample (240) → high population validity
* Repeated measures design → low particpant variability → high internal validity
* Free recall → high construct validity
* Multiple lists: some words might be easier or harder to rememeber depends on participant emotional attach
* Practice effect → affect the result

Limitations

* The testing method (repeated word) is quite simple compared to the complexity of the memory → hard to generate to real life → low external validity → not a clear indicator for clear separation of the STM and LTM
* Participants are all male → no diversity → low population validity
* Boredom, fatigue can impact the result
* Only 30 seconds → duration of STM can be longer → construct validity