Covert Solutions to Computerized Mathematics Problems as a Function of Visibility of the Numbers and Rate of Digit Presentation

Abstract

This exploratory investigation concerned estimation of sums of single digits as a function of the visibility of the digits, and the interstimulus intervals between digit presentations. A standard Televideo computer terminal (CRT) was used to randomly display digits ranging from 1 to 5 such that 25 numerals were successively presented per set. 77 undergraduate volunteers were randomly assigned to either a “visible” group in which numerals appeared on the CRT screen or to a “nonvisible” group in which the screen light was dimmed so that the digits were not visible. The rate of digit presentation varied within each group per set of numerals as a function of 3 interstimulus intervals (1, 3, and 5 sec.). Comparisons were made between participants' estimates of the sums of these numbers versus the correct sums given as feedback by the computer. A 2 × 2 × 2 mixed split-plot factorial analysis of variance gave significant differences between interstimulus intervals and significant interactions between interstimulus intervals and estimated versus correct means. There were no significant differences due to “viewing” conditions per se. Scheffé multiple comparisons gave significant differences between estimated versus correct sums at the 1-sec. interval for both visible and nonvisible viewing conditions. There were no significant differences between estimates and correct sums at the 3-sec. or 5-sec. interstimulus intervals for visible or nonvisible viewing conditions. Results are interpreted in terms of matching-to-sample, retrieval interference, and threshold for interstimulus intervals.