Soci 304 1nd Edition Lecture 4Outline of Last Lecture I. Types of UCROutline of Current Lecture II. Theory, correlationsIII. Principles of causal analysisIV. Validity of URCCurrent LectureTheory: to make causal arguments about causes of crime.- Make statements about relationships between observable phenomena.- The key is that is can be falsified (meaning that it can be proven to be wrong)Correlation: things tend to vary systematically in relation to each other- Such as weight and height- Positive correlation: as one variable goes up the other variable goes up as well. It can mean the direction as well. It has a positive direction. - Represented by positive in causal diagrams- Negative Correlation: as one variable goes up the other decreases- Ex: the more miles on a car the less it is worth- Represented in the negative causal diagramsR= Pearson correlation coefficient- Shows the degree of association between variables- Varies between -1 to +2- R=1 is positive and it represents a perfect correlation- R=0 means no correlation- R= -1 is negative and represents a perfect negative correlation.- Most of the time, almost always there is never a “perfect” correlation. They are more around .2 or .3- Lets us understand the relationship between two variables These notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute. Correlations yields information on the relationship between variables Theories make causal statements, needs to establish causality.  We need a theoretical rationale to have a good reason to believe a relationship existsPrinciples of Causal Analysis:- Harsh punishment = cause of delinquency- A causes B- A = independent variable and B= dependent variable. The outcome variable3 Criteria’s of Causality: 1. A and B are statistically associated= relationship Correlation coefficient is r= - 1 to +1 R= .23 = moderate positive association R=-.15 = weaker negative association  R=1 = perfect association 2. A is causally prior to B (causal order or temporal priority) The idea that the IV comes before the DV. If you measure two things at the same time we don’t know what came first. We can’t make a causal statement when we measure two things at once.3. The association between A and B does not disappear when the effect of other variables causally prior to both if the original variables are removed. Meaning lack of spuriousnessor non spuriousness. We need to isolate the relationship to make causality.  Spuriousness:Harsh discipline -------------- delinquency High Crime Neighborhoods ** If we measure families in the same type of neighborhood that have harsh parenting styles and we find it correlates with delinquency we can confidently say Harsh discipline = delinquency. The point is we have to isolate the High crime neighborhood variable to be confident in our results because if we don’t then the results could read that high crime neighborhoods and harsh discipline causes delinquency. PolicyCausal theories in social sciences deal with probabilities. Harsh erratic discipline increases the probability of delinquency. It can inform us what policy we should invest in. If supported by research, this has policy implicationsTo decrease delinquency, then it may be useful to try to change parenting stylesIndirect Association: harsh disciple is linked to delinquency because it causes angry feelingsA (harsh discipline) B Delinquency C. Angry or defiant feelings Gove UCR vs. Victimization1. Focus is on clarifying what the UCR measuresa. Validity: measuring what we intended to measureb. Face Validity: from the outside of the experiment does it seem like it measures what it is supposed to measure. c. Convergent Validity: focus on correlations (r). Comparing what we get from UCR and victimization. If the two things correlate, we have confidence that we are measuring crime. 2. Less clear evidence for rape and assaulta. Rape r= .01 to .38b. Assault r= -.39 to -.59 negative association- Larceny: evidence is not strong - Homicide: UCR is accurate - Motor theft: highest correlation between those sources r= .70 to .91- Burglary: r= .60 to .81 The highest validated crimes are motor vehicle theft, robbery, and burglary and homicide.Self- report vs. Official - Early self-report data were not showing the same relationship between race and class and patterns of delinquency captured by official data.- The controversy: self-report data generally finds no difference in delinquent behaviors byclass or race but official data finds differences by these social correlates. o 2 potential reasons for the controversy: 1. Problems with the official data2. Problems with the self report data and the way it was being measured ( this was actually the problem)The problem was that self-report data did not account for very serious crime.We solve this by a better response scale; they found more stimulation in patters from self-reportdata to patterns with official dataBy breaking out the response scale form 3 or more times to list the number of times they were able to tap high frequency offending. Quasi- experimentHigh Poverty NeighborhoodsControl Section 8 Experiment groupStay the same Move to a potentially less impoverished neighborhoodAll of these people lived in high poverty neighborhoods. Experiment group were helped to move in LOW poverty place. They got the full intervention into moving. Findings: violent crime dropped if you move people from a high poverty neighborhood to a low poverty neighborhood. But, there was a slight increase of property