Test 1 Content Flashcards
1
Q
statistics
A
- set of mathematical procedures for summarizing and interpreting observations
2
Q
descriptive statistics
A
- ex. mean and variance
- used to summarize or describe a set of observation
3
Q
inferential statistics
A
- ex. t-test and ANOVA
- used to interpret and draw inferences about a set of observations
- you are making inferences about a population from a sample of observations
4
Q
measures of central tendency
A
- measures that reflect where on the scale the distribution is centred
- mean: average score
- median: middle score
- mode: most frequent score
5
Q
measures of variability
A
- measures that reflect how much the scores are spread out
- range: highest - lowest
- variance: average of how much each score deviates
- standard deviation: square root of variance
6
Q
deviation scores
A
- telling us how far away the raw score is from the mean of distribution (mean - raw score)
7
Q
random sampling
A
- ensuring everyone has an equal chance of being samples
- if a sample is representative of the population of interest, you have external validity, which is important when wanting to generalize info
8
Q
random assignment
A
- subjects are randomly placed into the different conditions/groups of the experiment
- this allows for internal validity as it reduces the likelihood that the groups differ in a critical way other than the difference in condition
9
Q
hypothesis testing
A
- looking at multiple sets of values and deciding whether the differences between them are due to small chance fluctuations or due to the independent variable.
- due to chance or it is actually real?
- the more variability there is in your sample, the more variability you will expect between samples
10
Q
standard error of the mean
A
- the standard deviation of the sampling distribution of the statistic.. a measure of how stable we expect the statistic to be
- how much variation would you expect to find if you took repeated samples from the same population?
11
Q
central limit theorem
A
- if you repeatedly draw samples, compute means and the mean forms a distribution, as N increases, the distribution would become normal
- if there is variability within your sample, then you would expected to have variability between your samples.
12
Q
variance sum law
A
- the variance of the differences between two independent variables is equal to the sum of their variances
13
Q
p-values
A
- looking at the t value and determining, based on the percentage, the likelihood that the t value will be in that range
- if larger than the selected p-value, that means that the t value is very unlikely to exist and you reject the null hypothesis
14
Q
dependent t-test
A
- the same participants give data on two measure
- ex. before and after treatment
- have to make another section for calculating the difference between the two measures
15
Q
pooling variance
A
- if we assume both populations are equal, then the average of the sample variance would be a better estimate when having unequal sample sizes
16
Q
effect size
A
- saying how big or important the data is
- how many SD apart are the means?
- use cohen d with already determine effect size numbers
17
Q
type 1 error
A
- finding somethiing that doesnt exist, just by accident
- normally is .05
18
Q
type 2 error
A
- failure to find a true effect when it truly exist
- directly linked to powero
19
Q
null hypothesis
A
- no realtionship between the varaibles in the population from which the sample is drawn from
20
Q
alternative hypothesis
A
- raltionship between the 2 variables exist, the IV does have an effect on the DV, the means of the populations are not equal
21
Q
power
A
- probability of correctly rejecting the null hypothesis if the null is indeed false
- depends on the overlap between the sampling distributions
- sampe size, effect size and alpha all affect power
- power = 1 - beta (type 2 error)
- increase N, close the gap of overlap between sample distributions (more accurate representation of the population)
22
Q
ANOVA
A
- see what is the probability that 2+ means were drawn from the same population
- no restriction on the number of means that are invovoled
- As the number of compariosn increase, the probabilty of type 1 error occuring also increases
