Research DesignAnalysis CRJ 3013
Popular in Course
Alea Nicole Vick
verified elite notetaker
Popular in Criminal Justice
This 15 page Class Notes was uploaded by Michael Prosacco on Thursday October 29, 2015. The Class Notes belongs to CRJ 3013 at University of Texas at San Antonio taught by Staff in Fall. Since its upload, it has received 11 views. For similar materials see /class/231427/crj-3013-university-of-texas-at-san-antonio in Criminal Justice at University of Texas at San Antonio.
Reviews for Research DesignAnalysis
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 10/29/15
Research Design and Analysis Ward CR 3013901 11111 Social Science Research Methods Formal definition Systematic procedure technique or mode of inquiry used by a disciple to acquire knowledge Informal Definition The study of how we know what we know Experimental Reality Things we believe that are real from our experience Agreement Reality Things we believe that are real because everyone else seems to think so Second Hand Knowledge Authority Gain from people we regard as experts Tradition Passed down Something true because it always has been Media Information presented by the media TV shows Movies Scientific Standards for Agreement Reality Empirical Support Evidence data Logical Support Theory Human Distortion of Reality Errors in Human Inquiry 1 Inaccurate Observation Observations that fail to represent reality perceptions involve interpretations vaseface picture Fabricated Research Scatter Plots 2 Overgeneralization Observe only a few cases and assume what is true for them is true for all cases 3 Selective Observation Observe cases that confirm our expectations amp ignore cases that would not concern our expectations 4 Premature Closure Decide one has enough observations amp excludes additional info operates with amp reinforces overgeneralization and selective observation 5 Illogical Reasoning Ideas that are logically incorrect llthe exception that proves the rulequot 6 Ideology amp Politics Bias in defining research problems and conducting research that terms from ideology amp or politics What can you do to avoid human distortion Be objective vigilant systematic and skeptical Disclose procedures What others can do to avoid human distortion Replication use the same research methods to answer the same research question Peer Review Findings are subject to a double blind peerreview process Importance of research methods Consumers of research Producers of research Research Design and Analysis Ward CR 3013901 11311 Paradigms Frame our reference or a model that organizes our view of something Positivism Theoretical paradigm that states there are observable social laws that guide human behavior Empirical Research Production of knowledge biased upon experience or systematic observation Types of Explanations deographic full complete explanation of a single case Normothetic Paisimonias explanation of related cases Pure vs Applied Science Pure acquisition of knowledge for the sake of science Applied acquisition of knowledge for the sake of solving a particular problem Purpose of research Descriptive Seeks to define amp describe social phenomena 2 Exploratory Seeks to 39 39amp39 390 social I 39 of which little is known about 3 Explanatory Seeks to identify causes and effects of social phenomena seen as the main goal of social review 4 Evaluation applied Seeks to determine the overall effectiveness of a criminal justice intervention Considers both Causes amp effect and Implementation Inductive vs Deductive Inductive Specific General Deductive General Specific Attributes and Variables Attributes Characteristics that describe a person place or thing Variables Logical groups something that can vary Constants Takes only one value cannot vary Empirical Generalizations Identifying amp specifying general pattern s in the data Criminology Theory Systematic explanation for the observed facts amp laws that relate to a particular aspect of crime andor criminal behavior Theories must be falsifiable testable Levels of Theory Micro Explains why individuals engage in crime Research Design and Analysis Ward CR 3013901 Macro Explains why aggregates or groups of people engage in a crime Hypothesis Specific statements or predictions regarding the relationship between two variables IV amp DV IV Independent Variable Causes the dependant variable to change DV Dependent Variable One attempting to predict or change 111811 Ethics What is morally right amp wrong What is Ethics in Criminology amp Criminal Justice Behavior that conforms the standards of conduct criminology amp criminal justice researchers amp professional Ethics in conducting research 1 Voluntary General rule research participation should be voluntary amp subjects should be told what they are participating in lnvoluntary deception Complete deception Conducting research without subjects knowing they are participating in research Partial deception Subjects know they are participating in research but are fooled about some aspects of the research Debriefing Statements Statement following participation that lets subjects know in detailed info Discloses deception if any amp reason for research Discloses purpose 2 Avoidance of harm General rule Researchers should avoid causing subjects any physical or psychological harm Physical harms harm that occurs physically Psychological harms In the mind ie Trauma embarrassment When is harm avoidable RiskBenefit Ratio RiskBenefit ratio potential benefits of the knowledge gained from research that must outweigh the potential harms to the research subjects Benefits gt Risks 3 Protecting Subject Identity Anonymity vs Confidentiality Anonymity Subjects cannot be identified Confidentiality Can be identified however researcher takes precautions to ensure no one other than the researcher can identify How to promote compliance to ethical statements 1 Code of ethics Set of ethical principles that researchers willingly abide by to protect research subjects amp conduct research in an honest manor 2 Institutional Review Board Committees that review research proposals amp determine whether research is quotsafequot to carry out Research Design and Analysis Ward CR 3013901 Informed consent Documents Documents that fully inform subjects 1 Purpose 2 What the study is about 3 Compensation 4 RiskBenefits 5 ConfidentialityAnonymity 6 Voluntary participation Special populations that require additional protections a Prisoners b Juveniles Three special Conundrums 1 Mandatory reporting Researcher cannot keep a promise of confidentiality and must report data to authorizes 2 Withholding of treatment Denial of treatment to research subjects despite the fact that treatment has shown or is beginning to show utility 3 Research influences crime Conducting research facilitates andor displaces criminal activity Ethics in reporting research Researchers should be ethical when reporting research 1 Reveal limitations shortcoming of research are common place and should always e acknowledged Aside from ethical considerations revealing limitations can stimulate future research 2 Report negative findings Negative findings should be reported You should never play with data to find a statistically significant effect 3 Avoid staging findings Staging a finding occurs when a researcher first finds an effect and afterwards formulates the hypothesis to quotstagequot true empirical research 4 Protection of subject identity Protection of subjects identity pseudonyms i Fake names in reports and publications that protect the identity of subjects andor groups Major Ethical Controversies 1 Tuskegee Syphilis Study 2 Milgrams Obedience to Authority Study 3 Zimbardo s Stimulated Prison Study 012511 Necessary vs Sufficient Causes Necessary cause X must be present for Y to occur ie llyou must show up for the final to make an A in the coursequot X AY Final Sufficient Cause The presence of X essentially guarantees Y occurs Deterministic vs Probabilistic Causation Deterministic Causation X 9 Y X always causes Y ie Sufficient Cause Probabilistic Causation X 9 PY X increases the probability on occurring Research Design and Analysis Ward CR 3013901 Confusing Correlation with Causation llCorrelation does not imply causation Instead it is 1 of 3 requirements for causual inference When Can We Make a Causual Difference Three Requirements for causal inference 1 Empirical Association 1st requirement Two variables must be related to each other with each other ie They must vary together YCrime X Self Control scatter plot going downward How do we describe empirical associations Functional Form Linear Nonlinear Direction Positive variables vary in the same direction as X quot Y I Negative variables vary in opposite directions as X v Y I Strength Weak Moderate Strong Determined by correlation coefficient Ranges from 1 to 1 Negative values Negative association Positive values Positive association The stronger the association the closer to 1 in absolute value Why is Empirical Association Not Enough for Causation Association tells us that X and Y vary but it could be that X 9 Y or Y 9 X quot Appropriate timeorder 2quotd Requirement The IV must precede the DV in time X then YieT1X 9 TzY X Peers Y Crime 12711 Cross Sectional v Longitudinal Research Designs Cross Sectional All data are collected at a single point in time Longitudinal Data are collected at different points in time IV Data is collected at T1 DV data is collected at T2 Ensuring Appropriate timeorder Generally we use longitudinal research designs to ensure the requirement of timeorder is met however the requirement of time order can sometimes be inferred with crosssectional research designs Types of Longitudinal Research Designs Repeated crosssectional design aka trend studies Longitudinal study in which data are collected at two or more points in time from different samples of the same population Research Design and Analysis Ward CR 3013901 FixedSample panel design Longitudinal study in which data are collected from the same i individuals at two or more points in time Cohortbased design Longitudinal study in which data are collected from individuals in the same cohort at two or more points in time When CrossSectional designs may be sufficient for time order IV is fixed prior in variation to DV Retrospective Archival Data Cases are equivalent on DV prior to treatmentintervention Why are association amp timeorder both not enough for causation Association and timeorder tell us that X amp Y covary amp X comes before Y but it could be that Z a X amp Y to covary ie Kid Thinks X Seatbelt Sign Y Bumpy therefore X a Y But 2a X amp Y Z Turbulence Example of Spuriousness We have determined that poverty precedes crime and that the two are related ie Timeorder amp Association 3 NonSpuriousness 3rd requirement The relationship between X amp Y must not be caused by Z Validity The truthfulness of our casual inference is our conclusion that X increases the likelihood on actually true Validity Threats Reasons we may be incorrect in concluding that X caused Y 4 Types of Validity 1 Statistical Conclusion Validity The truthfulness that we have an empirical association between two variables as indicated by a statistical test Type 1 False Positive ie Pregnancy Test Type 2 False Negative 2 Internal Validity The truthfulness that we have a casual relationship not simply empirical association between two variables ie genetic influences Research Design and Analysis Ward CR 3013901 2111 3 External Validity The truthfulness that a causal relationship between two variables will hold in other contexts 4 Construct Validity The truthfulness that the observed causal relationship accurately represents the real underlying causal relationship Scientific Realism The study of how causal mechanisms operate in particular contexts quotquot39 39 Aquot quot 39 39 forthe 39 betweenXandY Context The social environment in which a causal relationship occurs Units of Observation v Units of Analysis U of O The level of social life or phenomena that is being observed U of A The level of social life or phenomena that is being studied Individuals Groups Organizations amp Social Artifacts Individuals Groups Families Census block groups police bears neighborhoods states Organizations quotthe state police agencies Social artifacts newspaper articles meeting notes neighborhood watch 20311 Poverty Absolute poverty see Putnam 1977 3x the S necessary for 2 months worth of food Relative Poverty Blau amp Blau 1992 Poverty is relative to a given place and a given time Subjective Poverty see Ruggles 1990 Minimum income of what someone thinks they need to make ends meet Conceptions Subjective mental images about the social world or other things we encounter Concepts Words or symbols in language that we use to represent similar conceptions Conceptualization The process of specifying exactly what a concept means Conceptual Definition A definition or specified meaning of a concept for the purposes of conducting research Dimension A specifiable aspect of a concept Research Design and Analysis Ward CR 3013901 INDIVIDUALS CONCEPTIONS 39 CONCEPT Low Self Control Some people cant control themselves Some people cant resist temptation Some people don t stop to think before they act v DIMENSIONS CONCEPTUAL DEFINITION Impulsivity Simple Tasks A A unidimensional personality trait Risk seeking physical activities self that predisposes persons to deviant centered temper behavior Concepts Observe or Measure We cannot observe concepts directly or indirectly however once we have a conceptual definition we can measure our concept Operationalization the process of specifying the procedure for actually measuring the concepts in our research Indicators What we use to measure a concept or a dimension of a concept From Conceptualization to Operationalization CONCEPTUALIZATION OPERATIONALIZATION Concept Dimension variable Indicator Income gt Monthly income SS Social Class Education gt Highest degree earned Occupational Prestige gt Job Type CEO scum owner manager supervisor worker pond scum Measurement The process of assigning numbers or labels to variables Measurement requirements exhaustive and mutually exclusive Exhaustive All possible attributes of a variable must be specified Mutually exclusive Each observation can take on only one attribute Continuous vs Discrete measures Continuous Varibables that can essentially have an infinite number of attributes Discrete Variables that have a fixed number of attributes Levels of measures The complexity of the mathematical means that can be used to express the relationship between attributes Research Design and Analysis Ward CR 3013901 Four levels Nominal Measures that simply tell us there is a different between attributes Race Ordinal Measures that rankorder differences between attributes Interval Measures that equally spaced intervals to meaningfully specify the difference between rankordered attributes Ratio Measures that use equally spaced intervals and a true zero point to meaningfully specify the difference and proportional difference between rankordered attributes respectively Dummy dichotomous measures Measures that are exhaustive with only 2 categories or attributes Can be nominal or ordinal Importance of levels of measurement for conducting research A given analytic techniques requires a certain minimum level of measurement Simplicity complexity predicament 21011 Does our measure suck Assessing measurement quality Reliability Whether a particular measurement technique will yield the same results each time Validity Whether a measure adequately reflects the meaning of a given concept Types of Reliability Assessments Testretest ntrarater Interrater Alternateforms Splithalf Internal consistency Assessing reliability Testretest An assessment of the similarity of a measurement at T1ie the test and T2 ie the retest O T Time ntrarater An assessment of the similarity of an observer39s coding at T1 and T2 Interrater I An assessment of the similarity of the coding of two or more observers Research Design and Analysis Ward CR 3013901 Alternative forms I An assessment of the similarity of subjects39 responses to two slightly different measures of a concept Splithalf I An assessment of the similarity of responses of 2 random groups each which are randomly given half the measures of the concept Internal consistency I An assessment of the similarity of responses to multiple itemes that are measuring the same concept I Cronbach39s alpha Types of Validity Face 0 An assessment of whether a measure appears to measure what it intents I Criterion 0 An assessment of whether results from a measure can compare to the results obtained from an established alternative measure ie the criterion I Content 0 An assessment of whether a measure covers the full range of a concept39s meaning I construct 0 An assessment of whether a measure is related to other measures as specified by a theory 21511 Scientific Realism The study of how causal mechanisms operate in particular contexts Context The social environment in which causal relationships occur External Validity The truthfulness that a causal relationship between two variables will hold in other contexts ReliabilityValidity Reliability is easier to establish Reliability is NOT validity You can get the same result over and over but it can be wrong Henry Margenau 1957 Measuring Crime Quote Research Design and Anaiysis Ward CR13013901 1 The trouble with the idea of measurement is be seeming clarity its obvious 132 its implied claim to nality in my inquieo my isc use Its status 39n philosop y at science is ca an m e utterly primitive hence the dif culties it embodies it any Land to escape detection and scrutiny Yet it cannot be primi u L a L ask which of two con icting measurements is Corral1 or preferable Forcibie RapeSTX v South Dakota with respect to forcibie rapes which state is more vi39oierit tai forcibie rapes in 2008 South Dakota 7 562 TeanrSOSS Rate per 100000 residents offercibie rapes in 2008 South D ota 7 699 4 Number oi Crimesvs Crime Rate Number of crimes Raw counts of crime ofCri39meS Crime Rate Counts of crime per 100000 persons Standardizing rates by 100000 persons is done by UCR Convention sta dardize by another number iimore appropriate Crime Rate is oi crimespopuiationi x100000 Main Sources for Crime Measurement Oi ciai Statistics Uniform Crime Report iUCRi Suppiementai Homicide Reports iSHRi Nationai incident Based Reporting System iNiBRSi victimization Surveys Nation ai Crime victimizati on Survey iNCVSi Seii Report Surveys Archivai Data of ciai Statistics UCR SHR amp NiBRS rirne statistics that are generated from poiice agency records Uniform Crime Report iUCRi History initiated by the iACP in 1929 0 in 1930 400 cities in 43 states participated Currentiy approximateiy 20000 agencies in aii 50 states participate 98 of urban iurisdictionsreport Over 90 of rurai Jurisdictions report Research Design and Analysis Ward CR 3013901 UCR Purpose To generate reliable statistics for law enforcement ie Operations management To assess crime trends over time within jurisdictions To provide a standard way to compare crime between jurisdictions UCR amp NIBRS How it works 1 Agencies determine If they will report Voluntary reporting except in some states 2 Agencies tabulate offenses Crime reporting handbook 3 Agencies send data to statelevel UCR program or directly to FBI 4 FBI releases report UCR Index of Crimes Part 1 offenses Murder and nonnegligent homicide Violent Crime index Red Robbery Property Crime Index Green Crime Index Red amp Green Aggravated assault Modi ed Crime Index Red Green Black Burglary Larceny Motor vehicle theft 0 UCR Crime Index and Modified Crime Arson added 1979 Index suspended in 2005 Crimes recorded from all victim reports and crimes observed by police UCR Indices UCR indices tell us trends in violent and property crimes Any problems with these indices How to lie How to lie with studying How to lie without studying Pa rt 2 offences Simple Assults Offenses against the family and children Disorderly conduct crimes recorded only if there is an arrest Supplemental Homicide Reports SHR UCR supplement that provides incidentlevel data on homicides Datelocation of offence Circumstances of crime Characteristics of victim and offender Victim offender relationship Research Design and Analysis Ward CR 3013901 Weapon use 21711 National IncidentBased Reporting System NIBR History In 1985 blueprint for the future if the Uniform Crime Reporting Program Three area of improvement Agencies to move to incidentbased reporting National UCR collect data on two levels limited and full participation National UCR would introduce quality assurance In 1989 first test submissions of NIBRS data In 2009 approx 44 of reporting agencies are certified for NIBRS 13 states submit all data via NIBRS NIBRS Purpose Seeks to provide more complete and useful crime data than UCR Incidentbased data NIBRS Incident Details 6 Segments group A required Administrative unique id date time Offense up to 10 separate offenses per incident Property each property loss is a separate segment Victim min of 1 incident per victim Offender Data on offenders Arrestee group B required One arrest record for all offenses linked Additional Info Added Biasmotivation 1990 Gang activity 1997 AssaultMurder of law enforcement officer 2003 Cargo theft 2005 UCRNIBRS Factors Affecting Reporting and thus recording Culture Seriousness of crime Victim of trauma Fear of retaliation Factors Affecting Recording Crackdowns ie Mandatory arrests Organizational changes ie New gang unit Officer discretion UCR Reporting Rules Arrest required for Pa rt 2 offenses Research Design and Analysis Ward CR 3013901 Hierarchy rule Data entry error UCR Limitations Many crimes omitted from counts Part 2 offenses require arrest Hierarchy rule quotDark figure of crime Summarybased measure NIBRS Limitations Some crimes omitted from counts Group B offenses require arrest quotDark figure of crime Current availability Key differences between NIBRS and UCR Incidentbased vs summarybased Group A 46 crimes22 offenses vs Part 1 8 crimes No hierarchy rule vs hierarchy rule 39 quot 39 vs no 439 I4 I I L L 22211 Victimization surveys NCVS Crime statistics that are generated by asking individuals about their victimization experience National Crime Victimization survey NCVS history Developed during the 1960 s and launched in 1973 as the National Crime Survey In 1993 updated and changed to NCVS New questions Improved methods Resulted in more crime reporting especially rapes and assaults Series break 1993 Recent Supplements Stalking 2006 School crime NCVS Purpose To develop detailed information about the victims and consequences of crime To estimated the number and types of crimes not reported to the police To provide uniform measures of selected types of crimes To permit comparisons over time and types of areas NVCS How it works Nationally representative sample Interview all persons in household 12 or older Fixedsample panel design Interview or telephone survey Screening questions Research Design and Analysis Ward CR 3013901 Retrospective reporting detailing incident NCVS offenses Rape Robbery Aggravated assault Simple assault Burglam Larceny l lII E ill lt Il39xefi NCVS Factors Affecting Reporting Participation in survey Memory Victim trauma Embarrassment Offendervictim relationship NCVS Limitations Many crimes are not included Crimes against children under 12 are not counted Selection bias National estimates only No state or local estimates Selfreport surveys SRS Crime statistics that are generated by asking individuals about their own offering behavior SRS history Selfreport surveys have been frequently used by criminologists to study violent crime and delinquency in general Examples Rochester Youth Development Study RYDS Boston Youth Study BYS collaboration with HYVPC SelfReport Surveys Limitations Accuracy of selfreport data Although for most purposes selfreports have been found to be reliable and valid Selection Bias Generalizability across different surveys Dark figure of crime Archival Data historical
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'