Measuring performance of immediate serial recall with the case of morphological influence

Xiaojian Li, Purdue University

Abstract

The goal of this study was to improve the measurement and modeling for immediate serial recall. An experiment was designed to examine the effect of an experimental factor of long-term memory, the morphological variation, on immediate memory. For both higher and lower word frequencies, two phonology-matched word pools were used to generate stimuli in order to uncomfound morphological variation from phonology and word frequency. Words in one pool were two-morpheme inflective words of some base verbs, nouns, and adjectives (e.g., BANNED, GUYS, LESSEN), while words in the pool of non-inflective homophones were not morphological variations of any word and had one morpheme (e.g., BAND, GUISE, LESSON). The method of the immediate serial recall was staircase. The influence of the experimental factor on memory span as well as on the reading time (rate) and recall time (rate) were examined. Two refinements of measuring recall performance and modeling immediate serial recall were made with strong mathematical support. First, sources of bias in some measures of span and recall time were discussed, and using computation formulas derived from this study, memory span was unbiasedly estimated and recall time was estimated to match the span length. The computation formulas were based on a finite Markov Chain model by which span was shown to be a limiting mean of stimulus levels instead of a mean of limiting distribution which was incorrectly assumed in other studies. Second, a covariance model treating the reading time (or rate) as a concomitant variable was adapted to specify the effect of long-term memory on immediate recall and to separate that effect on stimulus reading from immediate recall. The implication of the empirical findings to the mechanism by which morphological variation and word frequency influence immediate serial recall were discussed.

Degree

Ph.D.

Advisors

Schweickert, Purdue University.

Subject Area

Psychological tests|Cognitive therapy|Linguistics

COinS