Class Notes (811,155)
Canada (494,530)
Criminology (2,101)
CRIM 320 (52)

Crim 320 week 7.docx

4 Pages
Unlock Document

Simon Fraser University
CRIM 320
Patrick Lussier

1 Crim 320 Week 7 February 20, 2012 Last class before midterm MT: 2 hrs 15% of final grade 40 MC Lecture notes & the small book What to do before conducting a statistical analysis SCREENING DATA - Missing data - Univariate outliers - Normality - Data transformation A crucial step of quantitative research - If done properly it will - Increase power of statistical analysis (i.e., explained variance) - Minimize type I and type II errors - Minimize risk of biased estimation of population parameters Missing Data - Random missing data - Not attributable to the participant’s characteristics o Part of a questionnaire is lost o Data is incorrectly entered into the database o Changing an item on a questionnaire during the course of the study o Lost the contact information of a participant - Non random missing data - Attributable to the participants’ characteristics… o Cannot read, write, etc. o Does not want to answer a question bc feel the information is too sensitive o May think the information may be used against him/her later on o Cannot take part in the study because incarcerated, hospitalized or deceased 2 - Randomly or non-randomly missing? - Generally don’t know why the data is missing o If missing values are random, then not a validity threat o Not critical if <5% of a particular variable is missing (n>100) o Otherwise, might not be able to generalize results… o Example: MMPI and reading abilities  0 means absence of information, in Cambridge data set under missing, you can’t leave the 0 in your analysis, SPSS doesn’t know, if you’re running statistical analysis 0 will get analyzed as a value by spss  Patterns of missing data o Spss/analyze/missing value analysis/patterns  Determining whether missing (random?) o Recode the variable of interest into 2 categories  (Coded=0) all missing for which you have missing/unknown information  (coded=1) all other cases for which you have known values o For example  (coded=0) no information on CD of boy  (coded=1)information on CD of boy o Compare the two groups on all other variables of interest  Maternal attitude (overprotective, cruel, passive, etc.) o CHI-square, T-test, Anova etc o Determine whether your two groups are statistically different o Report those analyses in your study  Prepare data for statistical analyses o You may decide to drop a variable from your study  If too many missing values o You may decide to analyse only complete – case –  Remove all cases with a missing value  Especially if the –case- has missing values of more than 1 variable  Typical mistake = leaving scores of -0- which represent a missing value  Okay when few missing and at-random o You may decide to recode the missing value  Advanced methods  Imputation UNIVARIATE OUTLIERS - A case with an extreme value on one variable - Presence can be explained by: o Incorrect data entry
More Less

Related notes for CRIM 320

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.