SOCI 310

• What does independence refer to in terms of chi-square tests?

  Knowing one thing about 1 variable does not give you any extra info about the other

  2 separate pieces of information that does not affect or influence each other  
• What does dependence refer to in terms of chi-square tests?

  • When 1 variable does influence  the other
  • 1 is the cause and the other is the effect
  • The distribution of 1 variable changes according to the distribution of another variable
  • In the chi-square test, dependence between variables is assessed by examining the difference between the observed frequencies (the actual counts in the data) and the expected frequencies (the counts we would expect if the variables were independent). 
• What are contingency tables and why are they used? What kinds of patterns do they show?

  • Counts of cases across each level of each variable
  • A way to organize and display data when looking at multiple variables
  • associations or relationships between variables
  • Independence
  • Interaction effect-when the relationship between one variable and the dependent variable changes depending on the level of another affects
• What is the null hypothesis?

  Is the statement of "no difference" 

  The formation depends on the specific test being conducted 
• What does the researcher believe about the null hypothesis?

  That there is a difference between groups or that a relationship does exist and therefore wants to reject it
• What is the critical region?


  The area that contains the unlikely sample outcomes that we specify in advance and if they fall into this region we would reject the null hypothesis
• What would the outcome be if the sample outcome does not fall within the critical region?

  We would fail to reject the null hypothesis 
• How do we use appendix c?

   find df ;

  (1- # of rows) x (1-# of columns)

  Then use critical value 
• What is X2 (critical)?

  The score on the sampling distribution of all possible sample chi squares that marks the beginning of the critical region.
• What is X2 (obtained)?

  The test statistic as computed from sample results
• What is regression?

  • any kind of number
  • With each unit increase in X (independent variable), there is an increase (or decrease) in Y by (slope- dependent variable)
• What is correlation?

  A number between -1 and 1
• How is the strength of a correlation determined?

  • Close to 0 means no relationship
  • close to 1 means a strong positive relationship
  • close to -1 means a strong negative relationship
• What is the regression line?

  A single straight line that comes as close as possible to all data points
• Explain the components of the regression formula

  Y= a+ bx

  Y = Score on the dependent variable

  a = Y-intercept (where the regression line crosses the Y-axis)

  b = The slope of the regression line (amount of change produced in Y by a unit change in x) 

  x = score on the independent variable
• Before using the formation for the regression line, what must be calculated?

  a & b
• What is the interpretation of the slope (b)?

  The amount of change produced by the dependent variable (Y) by a unit change in the independent variable (X)

  Tells us how steep the line is and how much Y changes when X changes 
• What is the interpretation of the Y-intercept (a)?

  The intercept is equal to the value of Y when X is 0

  Tells us where the line starts on the Y-axis
• What is Pearson's r?

  A measure of association for 2 interval-ratio variables
• What is the strength of the relationships for Pearson's r?

  • Between 0.00-0.10 means its weak
  • Between 0.11-0.30 means its moderate
  • Greater than 0.30 means its strong
• What does a Pearson's r of 0.50 indicate?

  A strong positive linear relationship between the variables
• What is the coefficient of determination?

  The proportion of all variation in Y that is explained by x, found by squaring the value of Pearson's r
• What is multiple regression?

  A multivariate technique that breaks down the separate effects of the independent variables in the dependent variable

  Used to make predictions of the dependent variable
• Explain the components of the multiple regression equation

  Y = a + b1X1+ b2X2

  • a =the Y-intercept
  • b1 = the partial slope of the first independent variable (x1) on Y
  • b2 = the partial slope of the second independent variable (x2) on Y
• What is partial correlation?

  The correlation between 2 variables while controlling for 1 or more additional variables 

  Basically, the correlation thats "leftover" after ruling out the influence of the 2+ other independent variables
• What is control variables and how does it relate to partial correlation?

  • Is a third variable (z) that might affect a bivariate relationship
  • First, when comparing 2 variables, calculate a pearson's r value and determine their relationship
  • Then to determine if there is another variable that affects the relationship (partial correlation); calculate another correlation while controlling for the effect of that 3rd variable 
• What is a zero-order correlation?

  When the new correlation is different from the first 1 which suggests the 3rd variable is affecting the relationship between the 1st 2 variables 
• What are the steps for interpreting results fro multiple regression?

  1. Start with a simple bivariate regression model 
  2. Make a note of the bivariate slope (e.g, b1 in model 1: Y = a +bx)
  3.  Then add any additional; variables to the mode; that could explain this relationship (Model 2: Y = a + b1x1 + b2x2)
  4. How did the first slope change in model 2?
• How would you determine a direct relationship?

  MODEL 1

  1. Size of  b1 is large
  2. b1 is statistically significant

  MODEL 2

  1. X2 is included
  2. Size of b1 remains large
  3. b1 remains statistically significant 
• How would you determine a fully intervening or spuriousness relationship?

  MODEL 1

  1. Size of  b1 is large
  2. b1 is statistically significant

  MODEL 2

  1. X2 is included
  2. Size of b1 is now small 
  3. b1 is not statistically significant 
• How would you distinguish a fully intervening from a spuriousness relationship?

  Must rely on theoretical arguments
• How would you distinguish a partially intervening relationship?

  MODEL 1

  1. Size of  b1 is large
  2. b1 is statistically significant

  MODEL 2

  1. X2 is included
  2. Size of b1 gets smaller
  3. b1 may be statistically significant
• How would you determine a suppression relationship?

  MODEL 1

  1. Size of  b1 is small 
  2. b1 is  NOT statistically significant

  MODEL 2

  1. X2 is included
  2. Size of b1 is large
  3. b1 gets statistically significant 
• What is a direct (independent relationship)?

  The relationship between 2 variables ( x & y) does not change even after adding more variables ( X2, X3, X4, etc.) to the model
• What is a spurious relationship?

  The relationship between the first 2 variables disappears after taking a 3rd variable into account
• What is an intervening (mediating or indirect) relationship?

  One variable impacts another, which in turn impacts the outcome variable of interest
• What is a suppression relationship?

  A relationship between 2 variables (X1 & Y) appears only after adding another relevant variable (X2) to the model
• What is an interaction/ moderation relationship?

  Tells you the relationship between 2 variables depends upon what someone reports for the 2nd independent variable
• What are the steps for an interaction/ moderation relationship?

  1. Create a new variable by multiplying X1 and X2 for each observation
  2. Add the results to the model as a 3rd term (Y= a + b1x1 + b2x2 + b3x1x2)

  3. is b3 is large and statistically significant it means its an interaction effect
• What are the 3 types of measurement?

  1. Nominal 
  2. Ordinal
  3. inter-ratio