![]() ![]() I would say the relationship above is of moderate strength.įourth, are there outliers? We are particularly concerned about outliers that go against the general trend of the data, because these may exert a strong influence on our later measurements of association. ![]() Judging the strength of a relationship often takes practice. In practice, in the social sciences, we never expect our data to conform very closely to a straight line. On the other hand, if the points form a broad elliptical cloud, then we have a weak relationship. If all the points fall exactly on a straight line, then we have a very strong relationship. Third, what is the strength of the relationship. ![]() In this case, there is no evidence that the relationship is non-linear. This issue will become important later, because our two primary measures of association are based on the assumption of a linear relationship. If it has a curve to it, then the relationship would be non-linear. Second, is the relationship linear? I don’t mean here that the points fall exactly on a straight line (which is part of the next question) but rather does the general shape of the points appear to have any “curve” to it. States with higher median age tend to have lower violent crime rates. In the case above, it seems like we have a generally negative relationship. If y tends to be lower when x is higher and y tends to be higher when x is lower, then we have a negative relationship. On the other hand, if the variables move in opposite directions, then we have a negative relationship. if y tends to be higher when x is higher and y tends to be lower when x is lower, then we have a positive relationship. We refer to a relationship as positive if both variables move in the same direction. First, what is the direction of the relationship. What are we looking for when we look at a scatterplot? There are four important questions we can ask of the scatterplot. Thus, the selection of the dependent and independent variable is more about which way it more intuitively makes sense to interpret our results.įigure 24: Scatterplot of median age by the violent crime rate for all US states That association is the same regardless of which variable we set as the dependent and which we set as the independent. The language of dependent vs. independent variable is causal, but its important to remember that we are only measuring the association. We are interested in predicting life expectancy by income inequality, so the dependent variable is life expectancy and the independent variable is income inequality. For example, lets say we were interested in the relationship between income inequality and life expectancy. The independent variable is the variable that we treat as the predictor of the dependent variable. The dependent variable is the variable whose outcome we are interested in predicting. When examining the association between two quantitative variables, we usually distinguish the two variables by referring to one variable as the dependent variable and the other variable as the independent variable. Additionally, the major approach here of ordinary least squares regression turns out to be a very flexible, extendable method that we will build on later in the term. The techniques for looking at the association between two quantitative variables are more developed than the other two cases, so we will spend more time on this topic.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |