Download CJC 205 Exam 2: Measurement Quality, Reliability, Validity, and Experimental Design and more Exams Policy analysis in PDF only on Docsity!
CJC 205 EXAM ANS 2 WE WITHRS Q100%UESTIO VERINS AFIEDND!! CORRECT
criteria for measurement quality - ANSWER reliability and validity precise measurements are superior to imprecise ones but not always necessary reliability -repeatedly to the same thing, will yield the same results ANSWER matter of whether a particular measurement technique, applied
consistency and stability does not ensure accuracy issues with reliability -single observer ANSWER reliability is questioned every time there is only a
reliability is also questioned when there is more than one observer takingmeasurements
test-retest reliability -findings upon different administrators; an estimate of reliability obtained by correlating ANSWER degree to which test provides stable, consistent pairs of scores from the sample people on two different administrations of the same test interrater reliability - ANSWER agreement among raters on the measurement degree to which independent test administrators agree in their scoring of test data
validity -the concept under consideration ANSWER whether an empirical measure adequately reflects the meaning of
measuring what you say you are face validity - ANSWER the measure selected needs to make logical sense tells us degree to which a test looks like it measures what it's supposed to measure-reflects perception of appropriateness or obviousness of test
limitation: allows subject to manipulate results of measurement criterion-related validity -criterion ANSWER compare your measure with some external
a measure can be validated when it can be shown to predict scores on another measurethat is generally accepted as valid
multiple measures -concept ANSWER comparing your measure to other measures of the same
classical experiment - ANSWER independent and dependent variables pretesting and post-testing experimental and control groups randomization
confidence in statistical techniques and reasoning that rely on randomization criteria for causation - ANSWER temporal ordering correlationnon-spuriousness
temporal ordering - ANSWER the cause must come before the effect correlation - ANSWER one must happen because of the other non-spuriousness -explained by a third variable ANSWER relationship between the cause and effect could not be
minneapolis domestic violence experiment (1984) - ANSWER methods: arrest assailant - most effective send both parties to counselingsend assailant away for a couple of hours
results:
- 10% of arrested offenders re-assaulted victim
- 24% of those forced to leave re-offended- 19% of those referred to counseling re-offended in follow up period
conclusion:- best response was to arrest
problems:
- arresting could only make matters worse in some cases
- minneapolis is an affluent neighborhood kansas city preventative patrol experiment (1972-1973) - ANSWER internal threats to validity - ANSWER selection bias endogenous changehistory effect/external events contamination treatment misidentification selection bias -analysis in a way that randomization is not achieved ANSWER bias introduced by selected individuals, groups or data for
endures that sample obtained is not representative of intended population endogenous change - ANSWER maturation:
- from beginning of a study to the end, subjects mature and change naturally testing:
- repeatedly measuring participants leads to bias and them just getting better atmeasurements
statistical regression:- subjects who initially measure at an extreme are naturally inclined to measure closer to the mean
small sample size can be a problem issues with control of experimental conditions can impact statistical findings factorial design -independent variable and measures the dependent variable ANSWER experiment that consist of 2 or more levels of an
building blocks of classical experimental design -and control groups ANSWER number of experimental
number of and variation of independent variables number of pretest and posttest measurements procedures used to select subjects and assign them to a group quasi-experimental designs - ANSWER randomization is not possible quasi = "to a certain degree" threats to internal validity:
- not as controlled as experimental design two types:
- non-equivalent groups- time-series
non-equivalent groups - ANSWER randomization is not used to create equivalent
groups so it cannot be assumed the groups are equivalent select subjects for groups in a way that makes these groups as comparable as possible- matching
referred to as comparison groups rather than control groups cohort designs -same time ANSWER group of subjects who enter or leave an institution at the
cohort is exposed to some experimental stimulus time-series design - ANSWER longitudinal examining a series of observations on some variable over time
- examining trends in arrest rates for drunk driving over time interrupted time-series design
- used in cause-and-effect studies
- observations at t1 -> interruption -> observations at t simple interrupted time-series design -over a longer period of time before and after a treatment is introduced ANSWER making a number of observations
self-reports -commit criminal, delinquent, or deviant acts ANSWER for research attempting to explain or explore why people
- ex: offender surveys
- typically the focus is on the frequency of offending
surveys - what to do - ANSWER questions and statements add flexibility be clear short items make sure wording of questions matches intended audience surveys - what to avoid - ANSWER double-barreled questions
- asks 2 questions in 1 long questions - can be ambiguous negative items- example: drugs like marijuana should not be legal
biased items/terms- social desirability: people answer in a way that they think people want to hear -> avoid by making surveys confidential and anonymous issues with self-report surveys - ANSWER prisoners
- may be reluctant to admit their involvement in crime- may not accurately remember
- may lie solutions
- crime calendar
- asks on 2 separate occasions questionnaire format -avoid abbreviations ANSWER uncluttered don't put more than one question on a line contingency questions -they fall into a certain group ANSWER respondents are only asked to answer questions if
respondents may not have to answer every question matrix questions - ANSWER saves space in questionnaire respondents find it easier to complete makes it easier for respondents and researchers to compare and answer questions problem
- can generate a response set among some respondents- respondent boredom
ordering survey items - ANSWER start with the most interesting questions self-administered surveys -validity ANSWER high in reliability and generalizability, but weak in
easiest to complete and least expensive mail online
could get sent to email as junk survey: computer-assisted in-person interview - ANSWER pros: respondents may be more honest- not as concerned with confidentiality
- may produce high rates of reporting illegal behavior reduces potential for bias by the interviewer cons:computer/tablets are expensive; cost can add up reduces time at the end, but requires more up front work things could go wrong- batteries
- program complication surveys: telephone - ANSWER pros: time and money respondents may feel more comfortable giving sensitive answers to someone whenthey're not face-to-face
cons: people can hang up on you limited to people who have phones
- increasing number of cell phone rather than landlines randomly generated phone numbers- get random sample but numbers may no longer be in service
features of qualitative interviewing - ANSWER richness: thick descriptions that help lead to real understanding meaning: can shed light on complex processes that might otherwise go unnoticed shared cultural views: people living in similar communities or subcultures have shared cultural meanings thatare shared through communication
qualitative interview: structured interview - ANSWER predetermined questions standardized qualitative interview: semi-structured interview -semi-structured interview allow you to explore themes that emerge during the interview ANSWER still standardized, but
acceptable to use unscheduled probes- attention, clarification, continuation probe
qualitative interview: unstructured interview -interview guide ANSWER conversation or use of an free-flowing
recording data - ANSWER options: audio recording video recordingphotographs
after interview, you have to write up your field notes that were taken during and afterinterview
field research -examining processes over time ANSWER effective for studying the subtle nuances of behavior and for
strengths:
- depth of understanding it provides researchers
- flexibility- qualitative research is a great way to study active offenders in natural habitat
- validity weaknesses:
- issues with precision
- reliability: measurements can be personal solutions:- detailed instructions on how to classify observations
- reducing the degree of interpretation associated with observation grounded theory - ANSWER looking at notes and looking at similarities and differences seeing how people reacted and making sense of people's engagement
complete observer - ANSWER strictly watching and not interacting with targets full participant -in a setting ANSWER goes about ordinary life in a role or set of roles constructed
people see you as a participant and not a researcher pros: increases reliability and validity
- subjects may be more honest- natural setting
- don't know they are being studied cons:
- participation may affect what researcher is studying
- legal and physical risks- risks associated with setting
participant-as-observer -she is conducting research ANSWER participates with the group, but it is clear that he or
observer-as-participant -interacts with participants, but makes no pretense of being a participant ANSWER identifies himself or herself as a researcher and
example: going on a ride along with officers going native -that they become a part of it and lose sight of their objective ANSWER researcher becomes so involved with what they are studying
thinking units
- ways to make sense of all the stories that emerge from your interviews or observations examples:- meanings
- episodes
- relationships- encounters
- roles coding - ANSWER assigning units of meaning to your data creates categories for the variables that you analyze think of this process as organizing your dataopen coding selective coding helps you develop themes themes - ANSWER lower-level and higher-level concepts that emerge example: similarities and differences- example: work with deadly and non-deadly shooters
repetitions
- example: hearing something repeatedly in interviews
theory-related material
- example: how people solve problems... how police officers interpret threats secondary analysis -- surveys, field research, agency records, content analysis ANSWER includes all sources of data
others have collected data that other people can readily access and use become very popular
- primary data can be costly and time consuming sources: ICPSR, BJS, NACJD pros:
- cheap, fast, useful when it comes to evaluation cons:
- validity secondary data: agency records - ANSWER come from police, courts, corrections
- published statistics
- non-published agency records collected for internal use3. new data collected by agency staff for a specific research purpose
