Term
|
Definition
the extent to which a measure is repeatable or consistent. If it is reliable, repeating that same measure on the same case/observation will yield the same values for the variable
r=t/(t+e) |
|
|
Term
|
Definition
|
|
Term
|
Definition
test retest--perform the same test multiple times
alternative form--same test/concept, different implementation/method and different time, reliability assesed by correlation
Intercoder Reliability--different observer using the same measure |
|
|
Term
|
Definition
a valid measure accurately represents the concept it is trying to measure.
an invalid measure measures something else than what was intended |
|
|
Term
|
Definition
face validity,
content validity
construct validity |
|
|
Term
|
Definition
| seems like a valid way to measure a concept on the face value |
|
|
Term
|
Definition
| the amount to which a measure represents/addresses all faces of a concept |
|
|
Term
|
Definition
| the amount to which it is related to other measures that it is required to be related to. |
|
|
Term
|
Definition
| test retest--perform the same test multiple times |
|
|
Term
|
Definition
| alternative form--same test/concept, different implementation/method and different time, reliability assesed by correlation |
|
|
Term
|
Definition
| Intercoder Reliability--different observer using the same measure |
|
|
Term
|
Definition
1. clear goal
2. reliable and valid measure
3. comparaility
4. know your data |
|
|
Term
|
Definition
characterisits,
general and specific, declarative form, abt expected relationship, brief, direct, guided by a theory |
|
|
Term
|
Definition
measures symmetry,
not skewed: cases tend to cluster symmetrically around the mean and taper off evenly
skewed: one tail is longer and skinnier, affects mean not median |
|
|
Term
|
Definition
| right tail is longer and skinnier, pulls mean upward |
|
|
Term
|
Definition
| left tail is longer and skinnier, pulls mean downward |
|
|
Term
|
Definition
1. categorical
2. ordinal
3. continuous |
|
|
Term
|
Definition
| variables are clearly defined in seperate categories. cannot rank and there is no natural order |
|
|
Term
|
Definition
varaibles are clearly defined. you can rank. not equal unit differences
ex. political identification
left independent right |
|
|
Term
|
Definition
variables clearly defined, can rank, equal unit differences
ex. annual income |
|
|
Term
|
Definition
organize and summarize data
frequencies, porportions, percents, ratios, rates |
|
|
Term
|
Definition
| estimation of the center of a distribution of values, can use mean or median |
|
|
Term
|
Definition
| can be presented using a box and whisker plot |
|
|
Term
|
Definition
| how spread out a distribution is var(y)= s^2(y)={the sum of (yi- average y)^2} /[n-1] |
|
|
Term
|
Definition
average
(sum of Yi)/n
"expected value" |
|
|
Term
|
Definition
more accurate measure of the distribtion, measures how far an observation is from the mean. average difference between each of our cases and the mean
st dev= s(y)= square root of variance= square root of [(sum yi-mean y)^2/n-1] |
|
|
Term
|
Definition
sum (yi- mean 7)=0
the sum of (all variables of y minus the mean) is equal to zero |
|
|
Term
|
Definition
| the sum of (yi-mean y)^2 < the sum of (yi-c)^2 when c doesnt equal the mean of y |
|
|
Term
|
Definition
|
|
Term
|
Definition
| regardless of the distribution, the expected value(mean) for the sample will be normally distributed around the population expected value (mean) |
|
|
Term
|
Definition
skewness=0 kurtosis=0
mean=median=mode
1 st dev=68% of data
2 st dev=95% of data
3 st dev=99.7% of data |
|
|
Term
|
Definition
indicates the steepness of the statistical distribution
positive=very steep
negative=not steep, flatter |
|
|
Term
|
Definition
1. credible causal mechanism?
2. could y cause x?
3. covariation?
4. confounding variable z? |
|
|
Term
|
Definition
1.Don't let data drive the theory
2.consider only Empirical evidence
3.make your theories Causal
4.Avoid normative statements
5.Purse parsimony and generality |
|
|
Term
|
Definition
standard deviation of sample mean
sy/square root of n |
|
|
Term
| how to find the range of st deviation |
|
Definition
mean of y +/- 1 x se= range for 1 st dev(68%of data)
mean of y +/- 2 x se= range for 2 st dev(95%of data)
mean of y +/- 3 x se= range for 3 st dev(99.7%of data) |
|
|
