Ainley et al. (2002)

Preface
Ainley, M., Hidi, S., Berndorff, D. (2002). Interest, learning, and the psychological processes that mediate their relationship. Journal of Educational Psychology , 94 ( 3).

This article by Ainley, Hidi and Berndorff is interesting because it takes the basic constructs of Interest and Learning and their relationship to a new level by investigating the mediating processes, thus providing us with a richer understanding of the relationship.

Introduction
Ainley ranks interest with other motivational drivers such as goals and self-efficacy. Interest, in the sense of individual predisposition, Ainley refers to as Individual Interest, the first of three types of interest. Individual interest in a topic is expected to elicit positive affective reactions, lend persistence to the task, and to enhance learning. Individual interest in a topic may be a very specific interest or may be a very broad, general interest, or may lie somewhere between the two extremes. In this study, the level of interest in the topic was measured.

The second type of interest is Situational Interest, which is interest spawned by something in the environment, such as an item of human activity, or an outcome of the presentation. Situational interest, may elicit a broad range of affective reactions, some negative, some transient, that influence comprehension and recall. The 4 situations in this experiment were texts in areas of popular culture and science that were offered to the students to read.

The third type of interest is Topic Interest, which Ainley defines as interest triggered at the presentation of a specific topic, and contains elements of the other two types. In this study, the presentation was merely a very succinct title of a text. The combination of individual interest and situational interest aroused by human activities was expected to generate topic interest.

The research questions
How does individual interest contribute to topic interest;
How do situational factors (different texts) affect topic interest;
What processes mediate between topic interest and learning.

The experiment
The focus of this study was the topic interest exhibited by eigth- and ninth- grade students in each of 4 topics from social science books and culture: Chameleons, X-rays, Star Trek/ X-Files, and Body Image. There were 117 Australian (14.25 years mean) grade eight students, and 104 Canadian grade 9 students ( 14. 6 years mean), approximately 50% male, from lower middle- and middle-class schools.

Methodology
It might not have been Ainley’s primary intention, but a major feature of this experiment is the methodology. It was conducted almost entirely on computers, all the materials and the raw data collection. Measurements of interest were calculated in the program from the raw data.

With one exception, the entire experiment was accomplished using the Between the Lines computer program (Ainley, Hidi & Tran, 1997). Not only were the four texts onscreen, but some data were recorded automatically ( order of text access, time spent on text, how many parts of text accessed). Surveys were also answered onscreen by the students (feelings about the text, and a multiple-choice test following each reading).

Each student self-reported onscreen their level of interest twice before reading the text, and a third time, on paper, after completing the entire computer program. The first general interest survey covered individual interest in the five domains of personal health, animals, TV and Movies, Science, and Popular Music. The first four domains were related to the selected texts and the last one was included because of its reputation as a marker of adolescent development. The second survey before the readings was a measure of the students’ initial expected topic interests based only on terse titles of the readings, and on individual interests, if any.

Each of the four onscreen texts was approximately 750 words long and they were expected to be of balanced interest for this age group and for both genders, though not necessarily equally interesting. Each student was asked to read at least some of all four texts but could choose the order. Their choice of order was an indicator of situational interest and was recorded automatically.
Each text was divided into three segments. After reading as much as they wanted to of each segment, the student reported his affective response and intensity by selecting an indicative face (e.g. smiley face) from a panel of assorted faces. The student then could continue with the following segment or change texts. A student’s persistence score was calculated from the number of segments read and the time spent reading each segment. As soon as the student declared he was finished with a given text, he was given a 3-item onscreen comprehension test based on the segments read.

Following the complete set of computer-based exercises, each student was administered a paper questionnaire, the Two Factor Curiosity Scale, (Ainley, 1986), (alpha = .89, r=.90). One subscale measured the student’s general interest in learning. The other subscale measured a student’s prior knowledge of each topic.

Results
Individual interest for each student was calculated after a principal factors analysis of the reported knowledge and importance for each of the five domains ( personal health, animals, TV and Movies, science and popular music). The analysis reduced the 5*2=10 ratings to five, and rotation reduced it further to knowledge and importance in one domain. Affect was assessed as the first affect indicator chosen after reading the first segment of each text.

To find the contributions of five individual interest and general learning interest to levels of the four topic interests, MANCOVA was used, including gender as a between-subjects factor. General interest in learning, personal health and animals were found to be significant contributors to topic interest, for p < .05, with effect sizes in the medium range.

Among the correlations between the personal characteristics and resulting variables in the secondary process was the relation of gender to test scores, where girls scored significantly higher than boys only for the Body Image topic. General interest in learning was not significantly related to test scores for any topic.

To study the relationships between individual interests, topic interests and the mediating processes between interest and learning, a multivariate modeling procedure (Lisrel) was used that produced a two-level hierarchy of processes. The secondary process level was the influence of the personal characteristics on topic interest, as well as the influence of the same personal characteristics on the responses that followed the reading ( affect, persistence, test score). The primary process level was the relations between topic interest and the same responses. The results were combined and summarized in a set of path diagrams, one for each text.

Discussion
The computer-based program that Ainley developed was a brilliant idea. She was able to obtain data automatically in most cases, but obtained personal data in near real-time without serious disruption or interference. It was virtually a hands-off method. If this was an Experience Sampling Method, it was a better one.

Not surprisingly, one half of the students selected texts to read based on the order presented on screen, apparently exercising no discretion of selection. A method of index scoring, based on students’ choices, was a remedial method developed for this study.

Overall, this study was informative and technically well done.

Ainley et al. article download