Chapter 3 + 4 Flashcards
What is a discrete versus a continuous scale?
Discrete: has a sample space that can be counted like school grade= 1, 2 ,3 etc.
Continuous: a scale that includes decimals or fractions like height= 5.4 feet etc.
What are the 4 main types of scales and describe them
- Nominal
- a scale containing individual defined categories - Ordinal
- have an specific order that is clear and contain no zero point, ex. a scale where the possible choices are “never, sometimes, often and always” - Interval
- have meaningful differences between each number and have no zero point, ex. calendar years, distances - Ratio
- contains a true zero point where typically the values represent a magnitude
What is a negative skew versus a positive skew?
In a frequency distribution graph a positive skew starts off with high values on the x-axis and tail slowly tapers down, whereas in a negative skew, the tail tapers on the left and builds to higher scores among the x-axis
Describe kurtosis and its 3 variations
Kurtosis describes the steepness of a skew in its center
- Platykurtic= relatively flat
- Leptokurtic= relatively peaked
- mesokurtic= in the middle
When is pearson r used versus spearmen rho in correlation?
Pearson r= when both variables are relatively linearly related
spearmen rho= when n<30 and data is ordinal
Describe reliability versus validity
Reliability: Refers to how the tool of measurement should remain consistent, measuring in the same way each time
Validity: Whether a test actually measures what it purports to measure
What are the 4 main types of sampling?
- Stratified
- Random
- Purposive
- Incidental
What is measurement error ? Explain the formula
Measurement error is the uncertainty in any given observed score. Thus, the observed score is equal to the True score plus measurement error, X= T+E
What is systematic versus random error?
Systematic error influences a test in a particular and consistent direction while random error is unpredictable variations
What is variance in relation to error ?
The total observed variance of a test is the true score variance plus the error variance .
Thus we can understand reliability as the proportion of true variance that makes up total variance