{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

4.18 Andrews Research

# 4.18 Andrews Research - Limits of Statistical...

This preview shows page 1. Sign up to view the full content.

April 18, 2011 Andrews Research Word of the Day in situ: in the natural position or place Andrew’s Model for Academic Attendance Academic Attendance = B1 (proportion of blacks) + B2 (Alexander Coverage) + B3 (Freedom Summer Volunteers) + B4 (KKK Chapter) Reading a Regression Table-look for four things 1. Dependant variable 2. Independent variable 3. Regression coefficients and their signs 4. P-value (asterisk) (statistically significant What did Andrew find? Private academy attendance -varies positively with proportion of blacks among school-age children -varies positively with Alexander coverage -varies positively with presence of KKK chapter
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Limits of Statistical Methods-‘treatment’ is not usually randomly assigned-treatment and control groups could systematically differ in ways related to outcome of interest. -this limits our ability to say the ‘treatment’ is the cause of the effect Endogeneity: reverse causality; a common issue (Y causes X instead of X causing Y) Spurious effect: third variable is responsible for the association between others.-random sampling is not always possible-collection of reliable data is not always possible-may only be a few cases of phenomenon of interest (ex: revolution)...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online