The application of the first order system transfer function for fitting The California Verbal Learning Test Learning Curve

Abstract

Very few attempts have been made to apply a mathematical model to the learning curve in the California Verbal Learning Test list A immediate recall. Our rationale was to find out whether modeling of the learning curve can add additional information to the standard CVLT [corrected] measures. We applied a standard transfer function in the form Y = B3*exp(-B2*(X-1))+B4*(1-exp(-B2*(X-1))), where X is the trial number; Y is the number of recalled correct words, B2 is the learning rate, B3 is readiness to learn and B4 is ability to learn. The coefficients of the model were found to be independent measures not duplicating standard CVLT [corrected] measures. Regression analysis revealed that readiness to learn (B3) and ability to learn (B4) were significantly (p < .05) higher in a group of healthy participants than in a group of participants with type 2 diabetes mellitus (T2DM), but the learning rate (B2) did not differ (p > .2). The proposed model is appropriate for clinical application and as a guide for research and may be used as a good supplemental tool for the CVLT [corrected] and similar memory tests.

Assessment of the learning curve from the California Verbal Learning Test—Children's Version with the first-order system transfer function

“…Comparison of the model's coefficients with some standard CVLT measures was provided elsewhere (Stepanov et al, 2010). Because principles of the CVLT construction are the same in adult and children versions, we consider it to be possible to expand our treatment of the issue from the adult version (CVLT) to the children's version (CVLT-C).…”

“…The measures obtained from the regression analysis of the learning curve do Downloaded by [University of Tasmania] at 21:34 28 November 2014 not contradict standard CVLT measures. On the contrary, the model's coefficients provide additional important information, so that fitting the learning curve with the proposed model is appropriate for clinical application and as a guide for research and may be used as a good supplemental tool for the CVLT (Stepanov, Abramson, Wolf, & Convit, 2010).…”

Overall Memory Impairment Identification with Mathematical Modeling of the CVLT-II Learning Curve in Multiple Sclerosis

“…A discriminant function analysis of the data retained five measures: short delay free recall, recognition discrimination index, long delay free recall, semantic clustering, retroactive interference; with 70.0% of the participants correctly classified [16]. In addition to the conventional CVLT-II measures available from the test manual and its companion computer program, Stegen and coauthors [16] used learning measures derived from a mathematical model based on the first-order transfer function for the assessment of the learning curve [17, 18]. The model's measures were valid in combination with some other CVLT-II standard measures in discriminating MS patients from demographically matched controls.…”

“…Factors such as motivation [24], general health, and previous experience with the test might also influence the first trial recall. Thus, we treat B 3 as an estimator of the general functional state of a participant before starting the test and call it “readiness to learn” [18], hence B 3 is an abbreviation for “readiness to learn.” In our model, B 3 primarily represents attention span and short-term memory encoding process.…”

“…Other factors such as fluctuating emotional state, medication effects, pain symptoms, and effort also influence B 4 values [13]. We treat B 4 as an estimator of general ability to learn and call it “ability to learn” [18]; hence, B 4 is an abbreviation for “ability to learn”. In our model, B 4 is a measure of long-term memory.…”