Making sense of data PDF

Preview

Summary

Discusses how to analyze your data and apply results...

Description

Making Sense of Data! - Tables (frequency distributions, cross tabulations) ! Intuition behind cross tabs!

- Large-n data is especially useful for generating:! - Quantitative estimates of relationships between variables ! - Quantitative estimates of the uncertainty about those relationships ! - 2 key techniques for bivariate relationships:! - Cross tabulations! - Bivariate regresions ! - If we have multiple predictor we use multivariate regressions ! Crosstabs! - Support for the death penalty and gender in the US ! - There is not a statistically big enough difference between the two ! Terms! - y_i = obersved value of corruption for each country ! - Y-hat_i = alpha + Beta*x_i!

Notes on Research Design ! - Taking a random sample does not mean you can establish causation. For this you must randomly assign treatment …! - Be as specific as you can bout operationalization/measurement ! - Include question wordings/values for each question! - Not required, but you may consider consulting code books of existing surveys ! - World value survey example !

! How do we know these effects are significant?! - t-statistic = coefficient/standard error! - If this is greater than 2, this indicates significance at the 95% level ! - The coefficient tells us if there is a positive or negative relationship ! - What does “significant at the 95% level” mean again? 95% of the time the estimate you get will be between these two variables ! - We are quite confident that we are observing a real positive or negative effect (in other words, we can reject the null hypothesis that the true value of the coefficient is 0)! - If the t statistic is greater than the absolute value of 2 tells you if it is statistically significant! - if P-Value is less than .05 than it is in the 95% range, also tells you if its statistically significant !...