CHAPTER 8: DOING NONEXPERIMENTAL RESEARCH MAKING AND ASSESSING DIRECT OBSERVATIONS

Inter-rater reliability

The degree to which multiple observers agree in their classification or quantification of behavior.

Behavioral categories

The general and specific classes of behavior to be observed in a study.

Frequency Method

Counting how many times a behavior occurs.

Duration Method

Measuring how long a behavior lasts.

Intervals Method

Marking yes or no whether the behavior occurs within specified time intervals.

Time Sampling

Sampling a period of time to observe behaviors during that interval.

Individual Sampling

Assigning people at random to observe behaviors.

Event Sampling

Recording significant or rare events regardless of observer assignment.

Observation accuracy

Ensuring that observers are unbiased and trained to accurately record data.

Observer bias

Influencing observations with personal bias instead of objective criteria.

Quantitative data

Data collected represented numerically that can be analyzed statistically.

Qualitative data

Data that cannot be expressed numerically and is often descriptive.

Naturalistic observation

Observations made in a natural setting without interference.

Ethnography

A nonquantitative technique studying social interactions within a culture.

Participant observation

When researchers join and participate in the group being studied.

Nonparticipant observation

Observers watch without becoming part of the group activities.

Sociometry

Identifying and measuring interpersonal relationships within a group.

Case history

Intense study of a single case to understand its history and outcomes.

Archival research

Studying existing records and data rather than collecting new data.

Content analysis

Analyzing written or spoken records for specific occurrences.

Meta-analysis

Combining and comparing results from multiple studies to find overall patterns.

Cohen’s Kappa

A statistic used to assess interrater reliability by comparing observed and expected agreement.

Intraclass correlation coefficient

A measure of agreement between observers for numerical data.

Grounded theory

Developing theory based on systematically gathered and analyzed data.

Bias in observation interpretation

Challenges arising from the observer's interpretation affecting data accuracy.

Statistical procedures

Methods used in metanalysis to combine results from different studies.

Live observations

Making real-time observations to improve data accuracy.

Coding behaviors

Systematically categorizing observed behaviors for analysis.

Validity percent

Percent agreement necessary for reliability measures, aiming for 70% or higher.

Sociogram

A graphical representation of the interpersonal relationship choices within a group.

Information extraction

Using software to analyze large databases for specific information patterns.

Reliability correlation

Establishes if the order of observations is consistent across observers.

Training and pilot testing

Methods to reduce bias in observational research by preparing observers.

Observation ethics

Considering the ethical implications of participant involvement in research.

Research design

The overall strategy or structure used to conduct research in observational studies.

Data triangulation

Using multiple methods or sources to enhance the credibility of research findings.