Question 1

Suppose that the beta in a particular situation for predicting college grades from high school grades is .4 and that a person had a GPA that was 1 standard deviation above the mean (Z score = +1)in high school. What would the predicted score for that person's GPA in college be?

Accepted Answer

The answer of Suppose that the beta in a particular...

Question 2

In prediction with one predictor variable, the optimal number to use for beta (the standardized regression coefficient)is the:

Accepted Answer

The correlation coefficient (r) represents the strength and direction of the linear relationship between the predictor and the outcome variable. Beta (β) is the standardized regression coefficient and it shows the change in the outcome variable (in standard deviation units) for every one standard deviation increase in the predictor variable. Beta can be calculated by dividing the covariance between the predictor and outcome variables by the variance of the predictor variable. In simple linear regression with one predictor variable, the optimal value for beta is equal to the correlation coefficient between the predictor and outcome variables. Therefore, the best choice is B, the correlation coefficient.

Question 3

A person's predicted Z score on the criterion variable is found by multiplying that person's Z score on the predictor variable by a particular number called a:

Accepted Answer

The standardized regression coefficient, also known as beta, represents the change in standard deviation of the criterion variable for each one standard deviation change in the predictor variable. Multiplying the predictor's Z score by the beta produces the individual's predicted Z score on the criterion variable.

Question 4

In a true experiment, participants are __________ a particular level of a variable and then are measured on another variable.

Accepted Answer

In a true experiment, participants are randomly assigned to different levels of a variable, and then the researchers measure the effects of those different levels on another variable. This is what allows for causal inference to be made. Option A is incorrect because participants are not asked to comment about anything; rather, their behavior or responses are measured objectively. Option C is incorrect because whether a participant is considered significant or not has nothing to do with the experimental design. Option D is also incorrect because measuring a variable is not the same as assigning participants to different levels of a variable.

Question 5

The correlation coefficient is abbreviated by the letter r, which stands for:

Accepted Answer

The letter "r" in statistics stands for "regression," as it is used to denote the correlation coefficient, which measures the strength and direction of a linear relationship between two variables.

Question 6

The invention of the correlation coefficient occurred as part of research being conducted for which ultimate purpose?

Accepted Answer

The correlation coefficient was developed by Francis Galton as part of his research into eugenics, which aimed at improving the genetic quality of the human population through selective breeding.

Question 7

If researchers were to measure a couples' level of exciting activities at one time and then examine the quality of their marriages several years later (as done by Tsapelas, Aron, & Orbuch, 2009), they would be carrying out a:

Accepted Answer

This is a longitudinal study because it involves measuring a variable at one point in time and then examining a different variable at a later point in time to determine if there is a relationship between them.

Question 8

Suppose that you want to be able to predict college grades from information about high school grades. In this case, the college grades would be called the:

Accepted Answer

The criterion variable is the variable that is being predicted or explained by the predictor variable. In this case, the college grades are being predicted based on information about high school grades, making college grades the criterion variable.

Question 9

In the social sciences, a strategy to rule out at least one direction of causality is to conduct studies where people are measured:

Accepted Answer

The answer of In the social sciences, a strategy to...

Question 10

Suppose that you would like to predict a person's college GPA from the person's SAT score. In this situation, the SAT score is called the:

Accepted Answer

In a regression analysis where one variable (the predictor variable) is used to predict another variable (the criterion variable), the predictor variable is considered independent and the criterion variable is considered dependent. Therefore, the SAT score is the predictor variable and the college GPA is the criterion variable in this situation.

Question 11

Which of the following represents the strongest linear correlation?

Accepted Answer

The correlation coefficient of 0.92 represents a very strong linear correlation. The closer the correlation coefficient is to 1, the stronger the correlation. Thus, option C with a correlation coefficient of 0.92 is the strongest linear correlation.

Question 12

Researchers say that a correlation is significant if:

Accepted Answer

A significant correlation is one where there is a low probability (typically less than 5% or 1%) that the correlation occurred by chance alone. This is typically determined through statistical tests that calculate the probability of obtaining a correlation this big if there was no true correlation in the population at large. Option A is incorrect because a straight line on a scatter diagram does not necessarily indicate significance. Option C is incorrect because there is no universal cutoff for what constitutes a significant correlation; it depends on the specific study and sample size. Option D is incorrect because chi-square is not typically used to determine the significance of a correlation.

Question 13

Who is given credit for inventing the statistic called correlation?

Accepted Answer

Francis Galton is given credit for inventing correlation.

Question 14

Suppose you are making predictions about people's poetry ability based on scores on a general writing-ability test. However, you also happen to have information about their creativity scores when making your predictions. What is it called when you make a prediction using both pieces of information?

Accepted Answer

Multiple regression is the technique used to make predictions using two or more pieces of information. Ridge regression is a type of multiple regression that helps to alleviate multicollinearity problems. Bivariate regression is used when only one predictor variable is available. Proportion of variance is a measure of how much of the variability in a dependent variable can be explained by the independent variable(s).

Question 15

Two variables, X and Y, have a strong linear correlation. Under what conditions can you determine the direction of causality just from knowing the correlation coefficient?

Accepted Answer

The correlation coefficient, whether positive or negative, only measures the strength and direction of a linear relationship between two variables. It does not imply causation or the direction of causality. High power relates to the probability of detecting an effect if there is one, but also does not determine causality direction.

Question 16

Which of the following is a way for researchers to rule out alternative directions of causality?

Accepted Answer

A longitudinal study allows researchers to observe and measure changes in participants over time, which can help rule out alternative directions of causality. A true experiment involves randomly assigning participants to different conditions and manipulating the independent variable, which can also help rule out alternative explanations for the results. A cross-sectional study compares groups of participants at a single point in time and cannot rule out alternative directions of causality.

Question 17

A researcher notes that the correlation in the group she studies is so large that if there were not a correlation among people in general, there is less than a 5% chance of getting her correlation. Her correlation is thus:

Accepted Answer

A correlation that has less than a 5% chance of occurring by chance alone is considered statistically significant. Therefore, the researcher's correlation is significant.

Question 18

It is a good idea to make a scatter diagram before figuring the correlation coefficient. This is because:

Accepted Answer

Making a scatter diagram allows you to visually estimate the degree and direction of correlation, which can serve as a check on the calculations of the correlation coefficient. It is important to have an understanding of the relationship between the variables before calculating the correlation coefficient. The other answer choices are not true - a clear line does not negate the need for a correlation coefficient, there are not different procedures for positive or negative correlations, and a clear curvilinear correlation does not necessarily warrant figuring a correlation coefficient.

Question 19

Two variables, A and B, have a strong linear correlation. What is the direction of causality?

Accepted Answer

A strong linear correlation between two variables does not imply a specific direction of causality. It could mean that A causes B, B causes A, both are caused by a third variable, or they influence each other in a feedback loop.

Question 20

Place the four steps for computing the correlation coefficient into the correct order: 1. Add up the cross-products of the Z scores.
2) Change all scores to Z scores.
3) Divide by the number of people in the study.
4) Figure the cross-product of the Z scores for each person.

Accepted Answer

The correct order for computing the correlation coefficient starts with converting all scores to Z scores, then figuring the cross-product of the Z scores for each person, followed by adding up those cross-products, and finally dividing by the number of people in the study to get the correlation coefficient.

Quiz 3: Correlation and Prediction

Suppose that the beta in a particular situation for predicting college grades from high school grades is .4 and that a person had a GPA that was 1 standard deviation above the mean (Z score = +1) in high school. What would the predicted score for that person's GPA in college be?

In prediction with one predictor variable, the optimal number to use for beta (the standardized regression coefficient) is the: