NUR 80A/B Chapter Notes - Chapter 15 & 18: Multivariate Statistics, Analysis Of Covariance, Smoking And Pregnancy
Document Summary
Mulivariate staisics refers to analyses dealing with at least three but usually more variables simultaneously. Many nurse researchers use complex mulivariate staisics to analyze their data. Beter quality evidence but more challenging to understand research reports. Correlaions enable researchers to make predicions: because two variables are rarely perfectly correlated, ways to improve ability to predict dependent variable is by including more than one independent variable in the analysis. Might predict that infant birth weight is related to the amount of maternal-prenatal care. Collect data on birth weight and number of prenatal visits and then compute a correlaion coeicient to determine whether signiicant relaionship between the two variables exists (i. e. whether prenatal care would help predict infant birth weight) In muliple regression, the dependent variables are interval-level or raio-level variables. Independent variables (also called predictor variables in muliple regression) are either interval-level or raio-level variables or dichotomous nominal-level variables, such as male/female.