Developing a new measure for conceptual knowledge: The concept retrieval technique

The objective of this thesis is to propose a new measure of conceptual knowledge, that can be applied in diverse educational environments. The Concept Retrieval Technique will be proposed as a viable alternative to overcome the shortcomings of conventional assessment measures. The Concept Retrieval Technique requires students to freely recall concepts or ideas for a given topic with raters awarding one mark for each correctly recalled concept, when matched against a target word list. The retrieval mechanism in the Concept Retrieval Technique is based on spreading activation theory. Basically, this theory states that from one input concept, activation will spread along the links to other concepts that are connected to it. Then, from each of these concepts to connected others, and so on. The scores for the Concept Retrieval Technique are hypothesized to represent the knowledge network including relevant connections for an individual student. It is developed based on wellestablished research findings from cognitive psychology on how knowledge is organized in semantic networks of connected concepts. Studies have been conducted to establish the reliability and validity of the Concept Retrieval Technique as an instrument to measure conceptual knowledge. Finally, the Concept Retrieval Technique has been automated and the results have been analysed to maintain consistent inter-rater reliability with human raters to ensure the effectiveness of the machine-scoring process. Implications for educational measurement are discussed.

• View at Worldcat

Free Full Text ( Final Version , 30mb )