CCJS 320 FINAL EXAM STUDY GUIDE http quizlet com 12056707 ccjs300 final exam flash cards Unobtrusive Measures Archival Records Anything that is left behind or written down Secondary Analysis o Primary data sources our class data o Secondary data sources GSS Data Personal Documents Biographies o Electronic documents twitter etc Content Analysis Meta Analysis o The content of mass media o Focused on particular topic within a given field Observation Investigator attempts to examine the activity of respondents while keeping his her presence to a minimum All you re doing is observing Disguised Observation undercover research A situation that attempts to mimic or imitate key features of reality computer simulations jury studies Advantages of Unobtrusive Measures Keep yourself hidden stealth or Simulations o a Non reactivity 1 Respondents exhibit more natural behaviors 2 Avoid over reliance on attitudinal studies o b Using recording hardware software increases accuracy o c Collection of physical evidence is inconspicuous and gives respondents anonymity o d Use of archival records permits study of phenomena across time o e Observational studies firsthand information o f Nonreactive measures may be the best way to investigate some phenomena Disadvantages of Unobtrusive Measures o a Ethical Questions invasion of privacy how much is too much o b Respondents may be atypical depending upon who we observe and collect data from o c Time and Resources o d Observer Biases o e Issues of accuracy surviving archival records may be biased Validity Reliability 1 Thinking about ERROR in research What can go wrong Error is a kind of invalidity sources of error are always present A Validity asks the question Does my measuring instrument a question a set of questions an entire survey in fact measure what it claims to measure B Reliability involves the stability and consistency of measurement Reliability asks the question If the study were repeated would the instrument yield stable uniform measures 2 Research Myth busters How to spot bogus invalid claims in scientific research Robert Park 2003 A Discoverer pitches the claim directly to the media B The discoverer says that a powerful establishment is trying to suppress his her work C The scientific effect is always at the very limit of detection effect size is small D Evidence for a discovery is anecdotal Data is not the plural of anecdotal E The discoverer claims that a belief is credible because it has existed for centuries F The discoverer has worked in isolation G The discoverer must propose new laws of nature to explain an observation 3 Types of Validity a Face Validity Does the measuring instrument appear at face value to be measuring what I am attempting to measure Rosenberg Self Esteem Scale b Content Validity Look at each question and ask if it contributes to what you re measuring An item by item analysis c Construct Concept Validity Does it measure what you think you re measuring Looking for a correspondence between theoretical fit operational fit more theoretical and philosophical d Pragmatic Criterion Validity Can we use the instrument to get out of it what we want o 1 Concurrent How does this work now to allow us to measure what it is what we measure o 2 Predictive Can we use this when measuring thin gs in the future e Convergent Discriminant Validation Use multiple methods to measure multiple traits This is coming out of psychology should see two things o 1 Convergence of similar concepts across different measurement methods the method is not the most important thing what is more important is what you re measuring o 2 Discrimination of different concepts across the same measurement methods again the method is not the most important thing the content is and measuring accurately When using multiple methods of measurement the same or similar phenomena o Triangulation 4 Reliability Want stable and consistent replication of findings upon repeated measurements A Stable Respondent should give the same answer to the same question upon repeated measurements B Consistent Items used to measure same phenomena should be highly related or associated with each other 5 Types of Reliability A Test Retest Potential Problem Pretest bias present B Multiple different forms C Split Half Technique no testing effects 6 Reliability the statistic A Several different types available with SPSS Reliability procedure B Most common Cronbach s Alpha C Ranges from 0 to 1 00 similar to correlation coefficient no negative values Scaling Index Construction 1 Levels of Measurement ALL VARIABLES may be classified as belonging to a particular level of measurement a Nominal numbers assigned to the categories are merely labels Gender 1 male 2 female b Ordinal All properties of nominal and rank ordering of values of the variable Numbers imply some distance between the ranks Attitude Qs 1 strong 2 disagree 2 neutral etc c Interval All property of nominal ordinal and assume values of the variables equal intervals between the Temperature scales F and C IQ scores d Ratio All the properties of nominal ordinal interval levels and a true zero 0 point income twice as many or half as much 2 Scaling Procedures Using more than 1 question to measure complex concepts developing set of questions Why would we want to do this 1 Developing single question indicators of complex concepts may be difficult 2 Good for ordinal level variables Rosenberg Self Esteem Scale 3 Indexes Scales are efficient for data analysis several question scales give better Xbar is a very efficient estimator measures The mean 3 Distinction between scale and indexes Index Construct through simple accumulation of scores assigned to individual attributes Scale Constructed through the assignment of scores to patterns of attributes take advantage of any intensity structure that may exist amongst those attributes 4 How to devise multi question scales 1 All questions used should be relevant to the variable you are measuring 2 All questions used should be equally weighted 3 Try to use variables measured at the interval or ratio level of measurement 4 All questions need to be properly coded including reverse coded items 5 Arbitrary Scales Constructed by the researcher based primarily on face validity Ex Hagan Social Class Scale p 256 6 The UCR Uniform Crime Report specifically the UCR list of index offenses is an example of an arbitrary unweighted scale 7 Attitude scales 3 Major Types 1 Thurstone Scale based on expert judges selecting
View Full Document