Memory Decisions Research Articles

Differences in Memory Retrieval: A Construct Validity Investigation

Long-term memory retrieval efficiency was investigated as a potential underlying source of individual and developmental differences in cognitive functioning. Fourth-grade, eighth-grade, and college-aged subjects participated in a task using the Posner letter-matching paradigm. Letter pairs were presented simultaneously under physical-match and name-match instruction conditions. Reaction times were used to estimate parameters of long-term memory retrieval efficiency, basic encoding, decision, and response time, and name and physical output interference. Psychometric tests of verbal and spatial ability were included to assess convergent and discriminant validity of hypothesized relationships between aptitude test performance and basic cognitive processes. Developmental differences were observed in most but not all of the processing variables. Individual difference analyses indicated that less confounded estimates of processing parameters were not systematically related to verbal ability at any age level. Basic encoding and response speed was the most consistent correlate of spatial ability. The results suggest difficulties in previous interpretations of NIPI-verbal ability relationships. The study of cognitive processes in interaction and embedded in meaningful tasks is discussed.

Context effects in recognition memory: The frequency attribute

The first experiment determined whether frequency context would affect recognition memory decisions and frequency judgments. In the high-frequency context condition, 5 words were presented at study six times each prior to the section of the list containing the target items. In the low- frequency context condition, 30 words were presented at study one time each prior to the targets. The items tested were the same in the two conditions and were presented one, two, or three times each. Recognition performance and the judged frequency of target items presented once at study was higher in the high-frequency context condition than in the low-frequency context condition, but the opposite was true for items presented three times at study. The results of three subsequent recognition memory experiments suggested that encoding processes were critically involved.

The lengthening effect in sequential estimations of a short interval: an alternative explanation.

Experiments were carried out on the lengthening effect in sequential time estimations. It appeared that lengthening only occurred in the case of intuitive evaluation of an interval. Autocorrelation functions revealed that the estimation process may be described as a series of compensatory actions with respect to the last trial. It was assumed that the last estimate is more or less always experienced as too long or too short. In the resulting correction process the positive corrections reach higher mean values than the negative ones since the upper limit of the interval is not defined. In view of the negative autocorrelation in the last trial it was postulated that the correction process is not random, but is related to sequential response bias, i.e. too many alternations. Subsequently, several presentation and estimation modes were compared. The possibility that lengthening might result from trace decay of the standard stored in memory was doubtful since the phenomenon was also observed in the case of pure production on the basis of verbal instructions only. Apparently, it makes no difference whether a standard is presented or not. The distribution of the time estimates under various conditions revealed that the subject becomes more uncertain in proportion as the amount of external reference given increases, so that estimation process variables seem to be of greater importance than stimulus factors (standard presentation). It was assumed that the temporal memory could be described as a register in which only one event (the last) remains available and forms the subject's frame of reference. Since lengthening appeared to be connected with several cognitive variables of a general nature it was postulated that the effect is not only inherent in time estimation, but may be considered as a phenomenon occurring in repeated estimations of various stimuli as well. This assumption appeared to be correct, as analogous effects were observed in the case of repeated line drawing. It was shown that the increase of the time estimates could be influenced significantly by introducing different risk levels, i.e. the degree of lengthening is a function of the subject's decision criterion. Furthermore, there was some correlation between the compensatory actions and sequential response bias in a randomization task. It is concluded that the lengthening effect does not only show a general character, but that it might be described and explained adequately within the framework of the cognitive theory of time experience in which memory mechanisms, decision theory, and sequential response bias play a significant role.

Decision and decay processes in the short-term memory for length.

Decision and decay processes in the short-term memory for length.

Memory Limitation and Multistage Decision Processes

Sequential decision models have heretofore assumed a full memory decision maker. That is, the model is permitted to retain, to any degree of precision, all information needed to optimize decision performance. This information may include functions or variables that change with observations and thus often implies a decision maker which possesses a large amount of soft (erasable) memory. In simple multistage decision problems soft memory can be reduced to two variables?the log-odds ratio L and the available number of observations n. The log-odds ratio is a quantitative measure of the decision maker's opinion of the cause of the observed variate. This paper examines the effect of limiting the decision maker's soft memory by specifying an m-bit register for the random variable L. The theory for limited memory multistage decision processes is presented in which there are two simple hypotheses. Numerical results indicate that the 3-bit memory is, for practical purposes, equivalent to a full memory decision maker.