Flashcards
Psych Stats exam 2
Cards in this deck (43)
- 1LECTURE 7CORRELATION AND LINEAR REGRESSION
- 2What are 3 ways to measure relationships (in correlation)?1) scatterplots
2) covariance
3) Pearson's correlation coefficient - 31) Define correlation 2) what does it measure 3) is it typically linear or non-linear?1) A way of measuring the extent to which 2 variables are related
2) It measures the pattern of responses across variables
3) Typically LINEAR - 4Define regressionIt is used to identify the line that
best describes a linear relationship - 5Define slopeThe slope of the line describes the linear relationship between variables;
it can change if the variables change - 6Define interceptThe point at which a line intersects the Y
axis when X=0;
can be negative, positive, or zero - 71) Define Pearson's correlation coefficient 2) what is the range? 3) what does 0 mean1) Statistics that determines the extent of
the linear relationship between two
variables
2) ranges from -1 to +1
3) zero means NO correlation - 8In Pearson's correlation coefficient, 1) what does the sign indicate? 2) what does the number indicate?1) SIGN indicates DIRECTION
2) NUMBER indicates MAGNITUDE - 91) Define covariance 2) what do we look at when judging covariance? 2)1) Covariance is used to see if how the variables interact with each other, meaning if one increases, does the other increase or decrease or stay same?
2) look out for how much the score deviates from the mean; if they deviate from the mean similarly, then they are prolly related - 10the correlation coefficient is the ______ version of covariancestandardized
- 11Standardizing covariance solves what issue?Covariance is dependent on units of measurement (so, you can get 2 diff numeric values depending on the unit of measurement like miles vs kilometers);
standardizing it makes it a correlation coefficient - 12How do you mathematically standardize covariance?divide by the standard deviations of both variables
- 13The covariance is the average ___-___ _____cross-product deviations
- 14Define linear regressionA way of predicting the value of one
variable from another variable - 15The regression line is only a model based on the _____ ____existing data
- 16LECTURE 8CORRELATION & HYPOTHESIS TESTING
- 17What are 3 problems in correlation?1) non-linear relationships
2) restricted range
3) outliers - 18Step 1 of Hypothesis Testing:State null (H0: rho = 0) and alternative (H1: rho =/ 0) hypotheses
- 19Step 2 of Hypothesis Testing:Set criteria for a decision (expected value)
- alpha
- directionality
- degrees of freedom
- look at Table to find level of significance for "r" - 20Step 3 of Hypothesis Testing:compute the test statistic to calculate "r";
AKA use Pearson's correlation coefficient formula to find "r" - 21Step 4 of Hypothesis Testing:
- 22To report results for hypothesis testing, phrase it like this:1) There was a significant positive relationship
2) between the number of ads watched and the number of items bought
3) [r(3)=0.87,
p<0.05],
4) suggesting that people who
watched more ads also bought more items. - 23After finding "r", then what does r^2 (AKA "r squared") give you?r^2 = coefficient of determination
- 24When calculating Pearson's correlation coefficient, ASSUME these things:• The measured variables are at least on an interval scale
• The sample is independently and randomly selected from the population
• The distribution of Y scores is normal in the population at every value of X (bivariate
normality)
• The variance of Y scores is equal at every
value of X in the population - 25LECTURE 9PROBABILITY
- 26_______ _______ is based on probabilityinferential statistics
- 27To calculate probability, we need to know these 2 things:1) number of total possible outcomes, or the sample space
2) how often an outcome of interest occurs - 28What 2 rules can verify your answer when solving a probability equation?1) probability varies between 0 and 1 (closer to 1, the more probable)
2) probability can never be negative - 29Define marginal frequenciesMarginal frequencies are sums of the frequencies in the corresponding rows or columns (of a cell table)
- 30Define mutually exclusivewhen 2 outcomes CANNOT occur together
- 31Define exhaustiveset of outcomes that includes ALL POSSIBLE outcomes that could potentially occur
- 32What is (always) the probability of a set of outcomes that is both mutually exclusive & exhaustive?ALWAYS 1 (one)
- 33Define conditional probability & exampleThe likelihood of an event given some
other event (one event in relation to another)
ex: passing 2nd exam GIVEN THAT you passed 1st one - 34Define joint probability & exampleThe likelihood of each of two events
occurring (both events occur)
ex: passing 1st exam and 2nd exam - 35Define adding probabilities & exampleThe likelihood of at least one of two
events occurring
ex: 2 events could occur-
1) pass either 1st or 2nd exam
2) pass both exams - 36What is Bayes' theorem?takes the idea of "relationships among probability" and says you can alter that
- 37Sampling WITH replacement vs. Sampling WITHOUT replacementWITH: select something and put it back in the pool it came from, and then select something AGAIN
(ex: take a card out of a deck, and put it back)
WITHOUT: select something and DO NOT return it to the pool it came from, and then select something AGAIN
(ex: take a card out of a deck and hold on to it) - 38Define permutationAn ORDERED sequence of a set of objects
or events - 39Define combinationsA sequence of a set of objects or events
in which the ordering of elements is
IRRELEVANT - 40LECTURE 10NON-PARAMETRIC TESTING
- 41In what 3 instances are non-parametric tests used?1) to test hypotheses that do not make inferences about parameters in a population
2) to test hypotheses about data that can have any type of distribution
3) to analyze data on a nominal or ordinal scale of measurement - 42ww
- 43ww
Information
- Description
- flashcards of lectures as we learn them - help to prep for exam 2
- Created by
- Sophia Khan
- Views
- 7
- Rating