Week 1 Flashcards
(32 cards)
Why is data cleaning essential in research?
It ensures accurate statistical analysis by identifying and addressing missing data, outliers, and reliability concerns.
What are key reasons for cleaning data?
To ensure validity and reliability, handle missing data, confirm subscale reliability (e.g., Cronbach’s α), and meet statistical assumptions (normality, skewness, kurtosis, outliers).
How can missing data affect a study?
It can impact the validity of findings.
What is listwise deletion?
A method that removes entire cases with missing data.
What is casewise deletion?
A method that removes individual missing values but retains the rest of the case.
Problem with listwise/casewise deletion?
Reduces the population/sample size.
What is multiple imputation?
A statistical modeling approach that estimates and replaces missing data values.
What does MCAR stand for? Missing Completely At Random.
Missing Completely At Random.
How do we test if data is MCAR?
If p > .05 → data is MCAR → listwise deletion acceptable;
If p < .05 → data is not MCAR → imputation needed.
What does Cronbach’s α measure?
The internal reliability of a scale.
What Cronbach’s α value indicates acceptable reliability?
α > .70
What is an Independent Variable (IV)?
A variable manipulated or selected to observe its effect on another variable; represented by X.
What is a Dependent Variable (DV)?
A variable that is observed and measured; it changes in response to the IV; represented by Y.
What is internal validity?
The extent to which the change in Y can be attributed to X alone.
How can internal validity be improved?
Through stricter experimental control.
What is external validity?
The extent to which study findings (X–Y relationship) can be generalized to other settings.
How can external validity be improved?
By using more naturalistic or representative study conditions.
What is operationalisation?
The process of defining and measuring a concept in a precise way.
What is a nominal scale?
Categorical data with no order (e.g., eye color); no averages.
What is an ordinal scale?
Ordered categories (e.g., race rankings); meaningful order but not equal intervals.
What is an interval scale?
Numeric values with equal intervals but no true zero (e.g., Celsius temperature).
What is a ratio scale?
Numeric scale with a true zero, where multiplication/division are meaningful (e.g., response time).
