I. GENERAL ISSUES

A. One of the general issues explored within the field of human memory research is the manner in which memory is partitioned, with one of the main considerations in partitioning memory being whether memory should be divided into multiple systems or a single general memory.

B. Tulving (1985) has provided several reasons why a multiple-system approach is preferable over a single-system perspective.

Empirical evidence of dissociations leads to the conclusion that multiple systems, and not a unitary structure, are involved in the memory process.

1. A dissociation is said to occur when an experimental variable(s) has different effects on different tasks/measures.

2. One must be careful in interpreting dissociations, however, as task differences may arise from variations in task difficulty, implementation, or familiarity.

Tulving, Schacter, and Stark (1982)

IV#1 Type of test--word fragment completion or recognition

IV#2 Retention interval--1 hour or 7 days

 Phase I—learn a list of 96 words.

 Phase II (1 hr. later)—word-fragment completion test (  _gr_cul__re)
or a recognition test  (agriculture on the 1st list? yes or no)  on half of the study words.  

Phase III (7 days later)—tests were given for the other 48 study words. 

 For recognition test, performance declined over the retention interval.

 For the word completion test, performance remained the same across retention intervals

 Dissociation exists

II. COMPONENTS OF MEMORY APPROACH

A. One of the major approaches to partitioning memory allows for the division of memory into different storage types, with the primary types being short-term and long-term memory.

B. The connection(s) between short-term memory (STM) and long-term memory (LTM) has received a great deal of attention within memory research, and approaches studying these connections are referred to as dual-store, or modal models of memory.

1. Capacity

2. Forgetting

3.One of STM’s purposes is to transfer information into LTM.

C. Several dissociations lead to the conclusion that the partitioning of memory into STM and LTM is valid.

1. The occurrence of the serial position curve in recall tasks shows that memory is affected by different variables.

Primacy vs. recency effects

2. STM and LTM are also argued to be dissociated as a result of the patterns of memory loss observed in different amnesic patients.

D. Another organizational approach within memory research is to divide LTM into episodic and semantic divisions.

1. Episodic Memory (EM) is said to be our memory store for personal, or autobiographical, information.

2. Semantic Memory (SM) contains, more general, encyclopedic, knowledge that is not as personally significant as episodic memory.

3. However, attempts to dissociate EM and SM in laboratory settings have been difficult, leading some to question the validity of proposing separate episodic and semantic LTMs.

E. Another element said to exist within LTM is Procedural Learning abilities, which deal with the acquisition of perceptual, cognitive, and motor skills involved in HOW to do things, which is arguably distinct from knowing WHEN the skill was learned.

Tulving and Schacter (1990) have presented certain pieces of evidence showing that implicit and explicit memories are dissociable.

a. Performance on implicit and explicit tests have been found to be uncorrelated.

b. Research has shown that certain experimental manipulations influence explicit, but not implicit, memory.

c. Individual differences, such as age, have been found to influence explicit, but not implicit, retrieval of information.

d. Research on individuals with various types of brain damage shows that amnesics tend to have poor explicit abilities, yet their implicit memory skills are preserved, equal to those of non-brain damaged control subjects.

Warrington and Weiskrantz (1968) --amnesics generally show intact procedural memory but lack of declarative memory. Tower of Hanoi Example

III. STAGES OF MEMORY

A. An approach to memory that exists within the “components” issue is that which deals with the stages that information goes through as memories are formed.

B. The basic stages are encoding, storage, and retrieval, and research on these processes attempts to identify the effect of certain variables on the different stages.

C. Although the 3 stages cannot be totally isolated from one another, research has shown that experimental and neurological conditions can produce changes in isolated stages.

1. For example, being anxious during a test may not have much of an effect on your initial encoding or storage of information, but it can impair retrieval.

2. Also, patient H.M. can retrieve information that was encoded and stored prior to his brain damage, but has difficulty with the encoding of new information.

IV. PROCESSING APPROACHES --Will cover in Chapters 9&10

A. Depth of Processing Theory argues that a single LTM system exists, and the different memory effects observed within this system stem from the type of quality of information processing that occurs.

1. Maintenance Rehearsal involves the continuous, passive repetition of to-be-learned material.

2. Elaborative Rehearsal involves engaging in a meaningful and comprehension-based analysis of to-be-learned material.

B. Research on memory-impaired individuals, such as amnesics, shows that elaborative rehearsal does improve their memory ability, but deep processing does not eliminate their amnesia.

C. The notion of depth of processing has been criticized

1. There were no independent measures of “depth.”

2. The concept of “depth” was seen as too subjective.

D. The concept known as Transfer Appropriate Processing (TAP) also plays a role in how memories are processed.

1. TAP argues that in order for our memory abilities to function well, the cognitive conditions under which information was initially encoded must be reinstated at the time of retrieval.

2. TAP has also been used to explain the differences between explicit (conscious) and implicit (unconscious) memory.

V. CONNECTIONIST MODELS

A. Connectionist models of learning and memory attempt to simulate the neural events that underlie the acquisition, storage, and retrieval of information.

B. In trying to model learning and memory, connectionist networks have several core ideas.

1. Individual neurons are simulated as neural “units.”

2. Each unit can connect to many other units, allowing them to play a role in a variety of knowledge representations.

3. As pairs of units are continually activated together, the strength of the connection between them increases; the opposite is true of non-activated pairs.

4. The activation of any individual unit is typically dependent on receiving input from other units, a concept referred to as “threshold for activation.”