exam 2 wikistat jazz

Question 1

adjusted standardized residual

Answer 1

chi square research which shows the difference between the observed frequency (Fo) and expected frequency (Fe) for a cell, divided by standard error. ASR>1.96 is P

Question 2

cause in correlation

Answer 2

correlation shows the relation between two sets of scores 2 DVs. we can only determine cause with a true experiment, manipulating an IV and measuring its effect on the DV so we cannot normally draw cause and effect conclusions with correlation

Question 3

cell mean

Answer 3

the mean for any unique combination of the levels of the IV’s in a factorial design ANOVA. Cell means are what we graph and we might polish them to make an interaction stand out

Question 4

chi square test of independence

Answer 4

a nonparametric inferential test used when frequency data have been collected to determine how well an observed count of people over various categories fits some expected count. Do the counts depend on the categories people might fit in?

Question 5

coefficient of determination

Answer 5

the proportion of the variance in Y that is accounted for by X. This is our measure of strength with correlation we can; calculated by squaring the R (rxr) the correlation coefficient r2

Question 6

predictive confidence interval

Answer 6

an interval of a certain width that we feel confident (P

Question 7

correlation coefficient

Answer 7

a measure of the linear relationship between two sets of scores. It can be any decimal value between -1 (inverse) and +1 with a 0 indicating no linear relation and a 1 indicating a perfect linear relation

Question 8

correlational method

Answer 8

research method in which the relationship between at least two dependent variables is measured. Usually neither of the variables can be manipulated or controlled. Almost always both variables are dependent variables. When one of the variables is an independent variable then we have a quasi experimental design and perform regression analysis

Question 9

Cramers V

Answer 9

the standard effect size used with chi sqaure test of independence

Question 10

expected frequencies

Answer 10

the Fe in a contingency table if the sample represents the population. this idea shows up in chi square when SPSS gives Fe we can compare the Fo in our data against the Fe for the population. Expected Fe may come from theoretical or norm predictions based on other research of the population

Question 11

factorial design

Answer 11

a research design with more than one independent variable. This phrase comes up most when in reference to analysis of variance where a design and analysis involivng indepenet varaibles is called a factorial design and uses a factorial analys of variance

Question 12

interaction

Answer 12

the effect of each independent variable across levels of the other independent variable. this occurs in 2 way ANOVA; in the summary table the interaction line appears as IV1xIV2. when the means of each group are graphed, the interaction shows as non parallel lines. When writing about the interaction we have to say how the variables in combination influence the dependent variable

Question 13

intercept

Answer 13

the predicted value for Y when X is equal to 0 in a regression equation, or the point where the regression line crosses (or intercepts) the y axis. if the regression equation has the general linear form Y’ = a + bX (intercept is a)

Question 14

linear relation

Answer 14

a relation between two dependent variables best fit by a straight line. it is the relation deteceted by pearsons correlation and its represented on the scatter plot as the regression line. if we claim there is a correlation between two variables we mean a linear relation - no correlation is no linear relation could be nonlinear or curvilinear

Question 15

main effect

Answer 15

an effect of a single independent variable. this term shows up most often in analysis of variance. the result of a one way ANOVA is a simple main effect beacuse the study involves only 1 IV and the source table has only 1 F to reflect that main effect. in a 2way ANOVA there are two main effects and an interaction so there are 3 F values in the source table, two of which are main effects one for each IV

Question 16

marginal mean

Answer 16

the mean of each row and column in a table of the cell means for a 2 way ANOVA design. We use the marginal means to interpret main effects and we use them to polish the cell means and make the interaction stand out more cleaarly

Question 17

mixed design ANOVA

Answer 17

an inferential test used to analyze the data from a study with at least two IVs and 1 DV. at least 1 IV must be within groups or repeated measures variable and at least 1 IV must be a between groups variable. the computer output for a mixed design analysis is messy so we simplify by viewing sphericity assumed

Question 18

multiple regression

Answer 18

a procedure in correlation that uses 2 or more predictor variables in a regression equation. If there are two predictor variables a graph of the data is 3D - if there are more than two predictor variables the data can not easily be displayed graphically 2D

Question 19

negative corerlation

Answer 19

an inverse relation between two variables where an increase in one variable is related to a decrease in the other.

Question 20

nonlinear relation

Answer 20

a relation between variabes described by a line that breaks or curves in some way. if the scatter plot shows a disjoint pattern the r will be close to 0 (no linear relation)

Question 21

nonparametric statistic

Answer 21

a statistical test that does not involve any population parameters (mu / standard deviation) the underlying distribution may not be normal (chi square)

Question 22

observed frequencies

Answer 22

the number of participant in each category in our sample. symbolized Fo and chi square compares the Fo to Fe

Question 23

pearsons correlation

Answer 23

the most common of the correlational statistics when both variables are measured on the interval or ratio scale. the letter associated with pearsons correlation is r and these values are often reported in a correlation matrix, a table that shows the r values for all pairs of variables tested; pearsons r can range from 0 - 1 and can be negative or positive

Question 24

polishing the means

Answer 24

removing confounds from main effects or interactions. in ANOVA we often want to remove the effects of each variable in isolation from the interaction so the interaction stands out clearly. done by quantifying the main effects and subtracting them from the cell means so only the interaction of the 2 IVs are displayed in the cell means and the bar graph

Question 25

prediction

Answer 25

identifying the factors that allow us to predict when an event will occur done with correlation and regression equation

Question 26

regression analysis

Answer 26

allows us to predict Y based on knowing one or more other variables (x / z). regression analysis allows us to write the regreession equation and plot the regression line for that equation on the scatterplot. we can either predict directly from the regression equation or graphially by drawing prediction line on scatter plot

Question 27

regression line

Answer 27

also known as the BLUE line (best linear unbiased estimate) that comes closest to all the points

Question 28

regression to the mean

Answer 28

tendency of scores that are particulary high or low to drift toward the mean over time. the best prediction we can make of a persons score (Y) in a sample of scores is the mean, but regression analysis tells us how much better we can predict if we know X and the correlation between X and Y

Question 29

slope

Answer 29

if the r is positive b will also be positive and when r is negative so will b

Question 30

standard error of the estimate

Answer 30

a statistic that measures the typical distance between a regression line and the actual data points. SEE is used for calculating CI's; 95% CI:Yt = Y' +- 1.96(SEE

Question 31

third variable problem

Answer 31

the problem of cause in correlation. two variables may depend on another variable. this makes it difficult to determine causation with strictly correlational information. when variables A&B are correlated it may seem intuitively obvious that A causes B, but we cannot rule out the possibility that a third variable (C) is the cause of them both

Question 32

validity

Answer 32

the extent to which a test measures what it says it measures. if an outcome is not reliable it is probaby not valid in psychometrics we have many kinds of validty including; concurrent, criterion - related, predictive, convergent, congruent and consensual

Question 33

chi square is a test of

Answer 33

independence

Question 34

the proportion of variance in 1 variable that is related to the variance in another vairable is

Answer 34

coefficient of determination

Question 35

vertical axes of graph represents

Answer 35

a variable - independent variable

Question 36

r(6)=.97, p

Answer 36

the degrees of freedom

Question 37

Y'=8+5x, when x increases by 2 points Y' will increase by...

Answer 37

10 (x=3)

Question 38

in ANOVA what is another name we use for the independent variable?

Answer 38

factor

Question 39

what is mixed in a mixed design two way analysis of variance?

Answer 39

independent groups and repeated measure

Question 40

X2=5.99, V=41% what could you conclude about chi square?

Answer 40

significant with moderate dependence

Question 41

the appropriate way of citing chi square

Answer 41

X2(1,N=1227)=5.99, p < .05

Question 42

the value of "a" in the regression Y'=a+bX is

Answer 42

the intercept

Question 43

the value of "b" Y'=a+bX is

Answer 43

the slope

Question 44

multiple regression is

Answer 44

exactly what it sounds like dumby