Get ahead with a $300 test prep scholarship
| Enter to win by Tuesday 9/24
Stats Exam 2
Terms in this set (31)
A more abstract idea. Variables need to match with operational definition.
A definition of a variable in terms of the operations a researcher uses to measure or manipulate it.
There is only one variable.
Has two variables, nothing can be manipulated.
Can be manipulated.
Can be sorted into categories
Can be placed in a meaningful order
Has consistent spacing, but no meaningful zero point (equal spacing between units)
Has a meaningful zero point
The deviation score (the distance between a particular score and the mean) divided by standard deviation (average variability)
Positive z-scores are above the mean, negative are below
The deviance score (distance between sample mean, M, and population mean, mue) divided by standard error (average distance between the sample and population mean)
What does a z-test tell us?
What the likelihood is that a sample mean came from a particular population. (Are they the same?)
What is a sampling distribution?
A distribution of sample means from a population.
Standard error vs. Standard deviation
Standard error is the average distance of a sample mean from the center of a sampling distribution while standard deviation is the average difference between the scores in the distribution and the mean.
Null Hypothesis for One-Tailed
Group A <= Group B. Falsify the hypothesis
Alternative Hypothesis for One-Tailed
Group A > Group B
Null Hypothesis for Two-Tailed
Group A = Group B
Alternative Hypothesis for Two-Tailed
Group A =/ Group B
Confidence Interval Equation
The sample mean (M) +- Critical value (Za) times standard error
- If the mean falls within your interval, there is a chance greater than your alpha that the 2 groups are equal
- Gives us a way of estimating the range in which we expect the real value of a particular measurement to be
Effect Size Equation
The sample mean (M) minus the population mean (mue) divided by standard deviation.
- Top of equation: observed variability
- Bottom of equation: random variability
Interpretation Guidelines for Effect Size
- How likely something is to be the same and significant.
Type 1 Error
False positive - different
Type 2 Error
False negative - same
When are numbers significant?
alpha > p value
When are numbers not significant?
alpha <= p value
- Allow us to compare across different measures
- Allow us to translate a score from one measure to another measure
- Allow us to draw conclusions about the probability of getting certain scores
- Tells us how many standard deviations a score is from the mean
One-Tailed vs Two-Tailed
A one-tailed hypothesis specifies direction in the relationship between variables while a two-tailed only predicts if they are different.
The criterion we use to determine if we reject the null hypothesis.
- If something is less than this likelihood, we call them different.
The likelihood of getting these particular means when they're from the same group.
- If it is less than alpha, we conclude that there is a significant difference (we are confident that the groups are not the same)
- If it is greater than alpha, there is no significant difference.
Why do we use standard deviation for effect size?
We aren't trying to draw a conclusion about the probability of getting means but instead we are trying to get a sense of what the amount of variability is in our study.
In a two-tailed hypothesis why do we divide the alpha by the number of tails?
Half the alpha needs to be below the mean and half needs to be above the mean.