An evaluation of intact and embedded headings as schema cuing devices with non-narrative text

Abstract

This investigation employed a schema theory framework in examining the utility of intact and embedded headings as schema cuing devices with non-narrative text. The argument was advanced that the viability of the schema construct as an explanatory mechanism has been firmly established in typical prose processing paradigms but that these paradigms contain a number of shortcomings which have restricted the external validity of the findings. The principal shortcomings cited were: (a) the use of short, artificial prose as stimulus material, (b) focusing on schema for input rather than for output, (c) insufficient attention to individual differences, and (d) few, if any, attempts to train students to make more effective use of schematic cues; the present investigation sought to remedy these shortcomings. The specific objectives of the study were to examine the influence of headings (as schemata cuing devices), and training on the use of headings (both as input and output cues) with non-narrative text under conditions of immediate and delayed recall; both intact and embedded headings were employed in the investigation. Additionally, the relationships between certain individual differences and the experimental manipulations were examined as a supplementary objective. The results for the primary analyses (free recall-information scores as the dependent measure) indicated that input and output training on the use of headings was ineffective in comparison with students using their "normal" study methods. The results also indicated that students provided with text containing headings performed significantly better than students whose text did not contain these processing aids; the principal differences occurred in the delayed recall condition. Data on a number of supplementary dependent measures were also analyzed. These results indicated that: 1. The total amount of accurate information recalled {uncued and cued recall combined} was facilitated by incorporating intact and embedded headings into text; the major increases were observed at long term delays. Students receiving passages with the headings recalled approximately 17 percent more information at immediate recall and 41 percent more information at delayed recall than students whose passages did not contain the headings. 2. Students who received stimulus passages with headings produced better organized free recall responses than students whose passages did not contain the headings. 3. Students who received passages with headings had a better concept of the subordinate levels of the author's organization than students whose passages did not contain headings. Due to the relatively small amount of variance explained by the regression of the individual difference measures on the control group's performance scores, it was decided that meaningful statements regarding the relationships between these measures and the treatment manipulations could not be made. In general, the results of this study support the assumption that the presence of intact and embedded headings facilitates performance with non-narrative text, particularly at long term delays. The results tentatively suggest that these devices may be more useful as retrieval aids than as comprehension aids. The failure to find beneficial effects for training on the use of headings must be viewed as inconclusive due to the pragmatic constraints imposed on the present study.