Many environmental behavior models have been proposed to predict or explain environmental behavior in the past two decades. Due to limitations in practice, these models were mostly tested on a basis of self-reported measures of environmental behavior. How would these models perform if an observed measure of environmental behavior was used? What are the differences between the results of testing a model with a self-reported environmental behavior (SB) measure and the same model with an other-reported environmental behavior (OB) measure? And, what do these differences imply?
A major model, Ajzen¡¦s (1985) theory of planned behavior (TPB), was tested with SB and OB measures respectively with structural equation modeling (SEM) analysis. Data were collected through questionnaires and participant observation in which students living in dormitories in a university were trained to observe their 172 roommates¡¦ five environmental behaviors. The other-reports of these roommates¡¦ environmental behavior made by the observers were based on their observation as well as experiences in living with and understanding of their roommates.
Results showed that there existed significant frequency gaps between behavioral intention (BI) and SB and between SB and OB. The difference between SB and OB frequency was about 1.25 points on a 7-point scale. The BI-SB correlation (.87) was much larger than BI-OB correlation (.31) implying that SB was better predicted than OB was by BI. It was found that the subjects¡¦ SB tended to be conservative compared with BI and that OB was provided even more conservatively by the observers. BI could be the common cause of both SB and OB or SB could intervene in the relationship between BI and OB. A few significant differences were found in the results of testing the TPB model with SB and OB. The preference for the behavior might have played an essential role in predicting one¡¦s SB. The model had a very large proportion of variance explained (.814) because of the strong connection between BI and SB. It also seemed to have a better model fit when SB was used while the model¡¦s capability in predicting behavior was greatly reduced due to the large disparity between BI and OB when OB was used. Finally, it seemed that social desirability was a not cause of the SB-OB gap.
| Identifer | oai:union.ndltd.org:NSYSU/oai:NSYSU:etd-0427109-035706 |
| Date | 27 April 2009 |
| Creators | Chao, Yu-Long |
| Contributors | Ming-rea Kao, Kuen-Yung Jone, Chung-Ping Cheng, Pin-Yu Chu, San-Pui Lam, Shin-Cheng Yeh |
| Publisher | NSYSU |
| Source Sets | NSYSU Electronic Thesis and Dissertation Archive |
| Language | English |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0427109-035706 |
| Rights | withheld, Copyright information available at source archive |
Page generated in 0.0023 seconds