ERIC Number: ED028123
Record Type: RIE
Publication Date: 1969
Reference Count: 0
Establishing Reliability and Validity Estimates for Systematic Classroom Observation.
Webb, Jeaninne Nelson; Brown, Bob Burton
A study was designed to (1) compare two types of reliability in the observation of teachers' behavior, (2) explore the relationship between observer reliability and the validity of their systematic classroom observations, and (3) investigate the effects of training, observer beliefs, and the passage of time on reliability and validity estimates. Subjects were 32 experienced elementary school teachers, 16 of whom attended five 2-hour training sessions in the use of the Teacher Practices Observation Record (TPOR) and all of whom took the Personal Beliefs Inventory and the Teacher Practices Inventory to measure beliefs. Both groups viewed two films of classroom teacher behavior and used the TPOR to record observed behavior twice, once 10 weeks after training and again 10 weeks after that. Reliability coefficients were computed for between-observer agreement and within-observer agreement (the stability of an observer's response over time). These and criterion validity coefficients were used as responses in linear multiple regression analysis. Conclusions were that if observers share a common perceptual framework (as these did), between-observer agreement can be achieved easily with little or no training, but within-observer reliability is difficult to achieve. Therefore , training should focus on establishing the reliability of the individual observer. (JS)
Publication Type: N/A
Education Level: N/A
Authoring Institution: N/A
Identifiers: Personal Beliefs Inventory; Teacher Practices Inventory; Teacher Practices Observation Record
Note: Presented at the 1969 American Educational Research Association meeting, February 1969, Los Angeles, California.