TIPS FOR USING EXPLORIT
What
is the difference between an independent variable and a dependent variable?
When
using cross-tabulation, how can I tell which is the independent variable?
What is Cramer's V?
What is Pearson's R?
What does the term statistical
signficance mean?
How do I know if a relationship is statistically
significant?
Self Quiz
What is the difference between an independent variable and a dependent variable?
An independent variable is a variable that causes a change in the dependent variable. Variables are not independent or dependent on their own - it all depends on how they are used in a hypothesis. For example, if we hypothesize that drug usage increases dropout rates then drug usage would be our independent variable and dropout rates would be the dependent variable. We are saying that dropping out of school is dependent on using drugs. However, if we hypothesize that dropping out of school makes it more likely that someone will use drugs, then dropping out of school would be our independent variable and drug usage would be the dependent variable. We would be saying that whether or not someone uses drugs is dependent on whether or not they drop out of school. So, the same variables can be either dependent or independent depending on how the hypothesis is worded.
When using cross-tabulation, how can I tell which is the independent variable?
In cross-tabulation, the independent variable is always the column variable.
Cramer's V is called a measurement of association - it tells whether or not two variables are related to each other. Cramer's V ranges in value from 0 to 1.0, the higher the number the more strongly two variables are related to each other.
Pearson's R is another measure of association. Like Cramer's V, the higher the value of the number, the stronger the relationship between the two variables. However, unlike Cramer's V, Pearson's R can be either positive or negative. A positive correlation means that two variables move in the same direction - when one goes up the other goes up, and when one goes down the other goes down. For example, there is a positive correlation between income and education. Those with more education tend to earn more income and those with less education tend to earn less. A negative correlation means that two variables move in the opposite direction. For example, church attendance and divorce are negatively correlated. Those who go to church more often are less likely to get divorces, and those who go to church less often are more likely to get divorced.
What does the term statistical signficance mean?
When we say that a relationship is statistically significant that means that the mathematical relationship between two variables is strong enough that we can be fairly certain that the relationship is not a fluke. For example, suppose you were conducting a survey to see who is going to win an upcoming campus election. If you surveyed 200 CCSU students and found that 51 percent of your sample preferred Candidate A and 49 percent preferred Candidate B, you would probably be reluctant to state with a great deal of certainty that Candidate A is going to win the election. You couldn't be certain that your sample of 200 students represents the whole student body. However, if you were to increase the size of your sample to 1,000 students, or if you were to find that one candidate is twenty percentage points ahead rather than just 2, then you would have more confidence in making a prediction. When social scientists are 95 percent sure that there results are accurate, then they say that the relationship is statistically signficant. If a relationship is not statistically significant, then as far as social scientists are concerned, there is no demonstrable relationship between the two variables. For example, suppose a researcher hypothesizes that the DARE program will reduce drug usage among teenagers. When she conducts her research, she finds that those who graduated from DARE were less likely to use drugs, but the difference is not statistically significant. Statistically speaking, she would have to reject her hypothesis because she did not demonstrate that there was a relationship between the two variables. When you are examining a statistical relationship using Explorit, the only time that you can say that two variables are related is if the relationship is statistically significant.
How do I know if a relationship is statistically significant?
ExplorIt makes it easy for you. If either Cramer's V (cross-tabulation) or Pearsons R (scatterplots, correlation) is statistically significant there will be either one or two asterisks next to it. One asterisk means that the level of significance is .05 (which means that there is a 95 percent chance that the relationship is not a random fluke) and two asterisks means that the level of significance is .01 (which means that there is a 99 percent chance that the relationship is not a random fluke).
1. Which of the following values for Cramer's V signifies the
strongest relationship between two variables?
A. .250
B. .378
Answer: B - the higher the number the stronger the relationship.
2. A researcher finds that 18 percent of teens who smoke use
illegal drug compared to 15 percent of teens who don't smoke, but the difference
is not statistically significant. The researcher should report that based
on these results:
A. Teens who smoke are more likely to use illegal drugs.
B. Smoking probably leads to the use of illegal drugs.
C. There is not a relationship between smoking and illegal
drug use.
Answer: C - if the relationship is not statistically significant, then differences in percentages should be ignored.
3. Suppose there is a statistically significant negative
correlation between a community's median age and the crime rate. This means
that:
A. Communities with older residents have lower crime rates.
B. Communities with older residents have higher crime rates.
C. Median age and crime rates are not related.
Answer: A - a negative correlation means that an increase in one variable is related to a decrease in the other.
4. A researcher hypothesizes that government corruption
increases poverty rates. The independent variable is:
A. Poverty rates
B. Government corruption
Answer: B - poverty is dependent on the level of government corruption in this hypothesis.
5. A researcher hypothesizes that high poverty rates increase
government corruption. The independent variable is:
A. Poverty rates
B. Government corruption
Answer: A - government corruption is dependent on the poverty rate in this hypothesis
Please make an appointment to see me for assistance if you do not understand these concepts.