IR 211 Final Study Guide - External validity o The process of testing the validity of a measure, such as an index or scale, by examining its relationship to other, presumed indicators of the same variable  Ex: if the index really measures prejudice, it should correlate with other indicators of prejudice o Key problem is to have to worry about stimulus; when you try to generalize from what happened to the real world is a challenge – external invalidity o The Solomon four group design – addresses the problem of testing interaction with the stimulus; involves 4 groups of randomly assigned subjects; permits 4 meaningful comparisons - Internal validity o Experimentation is a method that works best when the tradeoff between external validity and internal validity is managed effectively o Internal invalidity occurs when experimental results may not accurately reflect what has gone on in the experiment itself; threat is present whenever anything other than the experimental stimulus can affect the dependent variable o 12 sources of internal invalidity  history – historical events may occur during the course of the experiment that may confound the results  maturation – of the participants [growing older – long term, growing tired – short term]  testing – process of testing and re-testing may influence ppls behavior  instrumentation – if different questionnaires, for ex, are use in pre and post testing, how can it be certain they are each measuring exactly the same thing? Issue of conceptualization and operationalization  statistical regression – danger that changes occurring by virtue of the participants starting out in extreme positions will be attributed erroneously to the effects of the experimental stimulus [ie extremely tall people are likely to have children shorter than them]  selection biases – groups need to be comparable at the start of an experiment  experimental mortality – subjects may drop out of an experiment before it is completed  causal time order – ambiguity about the time order of the experimental stimulus and the DV [may be the DV caused the changes in stimulus instead of vice versa]  diffusion or imitation of treatments – experimental group might pass on elements of the experiment to the control group and control group is “contaminated”  compensation – subjects in the control group are often deprived of something to be of value; they may be treated differently out of guilt – no longer a genuine control group  compensatory rivalry – subjects deprived of experimental stimulus may try to compensate by working harder  demoralization – feelings of deprivation within the control group may result in their giving up - Survey validity o If you ask people about something they know nothing about, you will get skewed answers because people don’t want to look ignorant o Bias and wording is a menace with surveys; need to seek neutral wording o Changes in what is acceptable socially can change very fast - Naturalism o Naturalism is the most time honored approach to field research – based on the assumption that an objective social reality exists and can be observed/reported accurately  Naturalism – belief in objective reality - Ex: UN can be observed and understood through our experiences IR 211 Patrick James 2012 Fall Weeks: 1 - 4 Midterm # 2 Study Guide Critics say that things that are really important are not the parts that are objective  Naturalists believe in nomothetic explanations – the experiences of others will be clear to me and vice and versa o Reasons for naturalistic observation:  To describe nature – leave variable uncontrolled and are of limited use for testing causal hypotheses, but descriptions can provide evidence that support or discredit hypotheses  To generate and test hypotheses – observations can lead to hypotheses  to assess behavior – changes in behavior can be measured and evaluated even if the starting point is unknown  Because other methods are impractical – ethical considerations limit some experiments; experiments can not answer all questions; experiments may lead to biased responses - Quantitative Data o Numerical representation of hypotheses, of our findings, and our arguments; mathematical statements about data o Not in opposition to qualitative, but rather complementary to o Often makes observation more explicit, easier to aggregate, compare and summarize data o Tends to be less rich in meaning than qualitative analysis but more straightforward; more precise information - Bivariate Analysis o The analysis of two variables simultaneously; purpose is descriptive; adds the element of determining the relationship between variables themselves – focus on the variables and their empirical relationships  Ex: church attended recorded by men and women – shows that women attended church more often than men; same table also suggests that the variable gender has an effect on the variable church attendance [no longer talking about men and women as separate subgroups but as gender as a variable] o Has been important in generating results and moving social science forward in the connections we see and in the things that don’t work out  Ex: population density, Malthusian [more density of people = greater conflict - interstate] – didn’t work; what does seem to have some connection is population density and civil warfare - Stratified sampling o Trying to get more representativeness by decreasing probably sampling error; group units into homogenous collections before sampling out of them o A method for obtaining a greater degree of representativeness by decreasing the probable sampling error [sampling error is reduced by increasing the sample size and by having a homogenous population] o Rather than selecting a sample from the total population at large, the researcher ensures that appropriate numbers of elements are drawn from homogenous subsets of that population  Ex: a stratified sample of university students – you would first organize your population by college class and draw appropriate numbers of freshman, sophomore, juniors, and seniors o Ultimate function is to organize population into homogenous subsets - Test Variable o Also called the control variable o The variable that is held constant in an attempt to clarify further the relationship between