Unobtrusive Measures 1 Definition nonreactive methods of gathering data means of obtaining information in which respondents are not aware of being studied o Idea came from originally a book titled Unobtrusive Measures Nonreactive Research in the Social Sciences Webb Campbell Schwartz Sechrest 1966 o Lack of awareness eliminates reactivity or stooge effects artificial behavior 2 Physical Trace Analysis the study of deposits accretion of matter and other indirect substances produced by previous human interaction physical trace evidence is not full proof o Forensics on Trial Episode from NOVA Oct 2012 o Example former President Clinton stain on Monica s dress This shows how much technology how improved and how quickly and accurately we can collect physical trace evidence 3 Archival Records existing data and autobiographies can contain much information that can provide a historical overview of criminological issues the way we would have access to this information is why its an unobtrusive measure o Secondary Analysis reanalysis of data that were previously gathered for other purposes Primary Data data we generated in class your own data raw data accompanied by any analysis Secondary Data GSS data basically any data that s out there analysis syntheses and evaluations o Personal Documents Social Media o Historical Data Potentially hazardous and relatively neglected area of the criminal justice system o Content Analysis where we are looking at the content of typically mass communication stuff written down videos Basic procedure in content analysis Selection of categories and subjects to be analyzed Rigorous establishment of criteria for inclusion a feature which ensures that the study can be replicated by others Carefully following the reestablished classification scheme Statistically analyzing the results Example if you are interested in studying attitudes toward capital punishment you could go back and look at blogs etc and you could see how attitudes have changed if at all o Meta Analysis quantitative analysis that reviews combines and summarizes the results of many differed studies dealing with the same research questions 4 Observation o Strategies in which the researcher s participation with the subjects is kept at a minimum and the investigator carefully records activities of the subjects o Strategy of data collection in which the investigator attempts to examine the activity of subjects while keeping his presence either secret or to a minimum 5 Disguised Observation researcher purposely hides his or her presence o Confederates persons planted to facilitate that which is to be observed speeds up the action 6 Simulations 7 Advantages o Non reactivity o Good way of avoiding problems with unobtrusive measures o A variety of gaming strategies that attempt to imitate a more complex social reality should require some sort of debriefing Exhibit more natural behaviors Avoid overreliance on attitudinal data LaPierre study o Using recording hardware software can increase accuracy of information o Collection of physical evidence mostly inconspicuous may afford anonymity o Using archival records permits study of phenomena across time o Observational studies Researcher gains first hand information o Non reactivity measures may be the best way to study some phenomena 8 Disadvantages o Ethical questions Right to confidentiality o Are the respondents typical or unrepresentative o Time Resources o Observer Coder Biases 1 Thinking about ERROR in our research what can go wrong o Error as a kind of invalidity how can we attempt to minimize error o Validity does my measuring instrument in fact measure what it claims to measure A question a set of questions an entire survey o Reliability questions the stability and consistency of measurement if the research or study were repeated would the instrument yield stable and uniform measures 2 Research mythbusters How to spot bogus invalid claims in scientific research Park 2003 o discoverer pitches the claim directly to the media o discoverer says that a powerful establishment is trying to suppress his work o the scientific effect involved is always at the very limit of detection o evidence for a discovery is anecdotal unreliable o the discoverer claims that a belief is credible because it has existed for Validity and Reliability centuries o the discoverer has worked in isolation o the discoverer must propose new laws of nature to explain an observation 3 Ways of determining validity types of validity o face validity does my measuring instrument appear at face value to be measuring what I m trying to measure Rosenberg Self Esteem Scale has high face validity Duck if it quacks like a duck then it is a duck o Content validity examine each question or item in a survey do the items measure the concept in question Item by item analysis designed to measure o Construct concept validity does the instrument measure what it has been A philosophical theoretical kind of validity What is the fit between the theoretical and operational of terms o Pragmatic criterion validity does the instrument work two subtypes Concurrent concerns itself with whether the measure enhances the ability to gage present characteristics of the item in question Predictive ability to accurately forecast future events of conditions o Convergent discriminate validity use multiple methods to measure multiple traits should observe a processer A convergence of similar concepts across different measurement A discrimination of different concepts across similar measurement methods AND methods o Triangulation the use of multiple methods to measure the same 1 Definition demonstrated through stable and consistent replication of findings on repeated measurements o stable respondents should give the same answer to the same questions on the retest o consistent are the items questions used to measure some phenomena highly related or associated with each other 2 Examine inter item correlations think about the Rosenberg self esteem scale 3 Types of Reliability o test retest the same instrument is administered twice to the same population and if the results are the same stability of measurement is assumed o multiple forms involve the administration of alternate forms of the instrument to the same group o split half technique does not involve retesting and substitutes the assessment of internal consistency of temporal consistency each half of a scale is analyzed as if it were a separate scale phenomenon