The Policy Process Model Source: See Bretschneider Module 5 Evaluation Approaches Tip: Need a logic model to evaluate well. p. 147 Rossi, Lipsey, and Freeman. (2004)/ Evaluation: A Systematic Approach. Sage. Adapted from United Way p. 95 Rossi, Lipsey, and Freeman. (2004)/ Evaluation: A Systematic Approach. Sage. p. 143 Rossi, Lipsey, and Freeman. (2004)/ Evaluation: A Systematic Approach. Sage. Program Evaluation Formative vs Summative Formative Main purpose: Evaluates how well pieces of logic model are fitting together (as you implement the program). If Summative Main Purpose: Assessing whether Short, Intermediate, and LogTerm Outcomes are Met a program fails, it could be because the theory is wrong (more education does not reduce drug use) or because it was implemented incorrectly (i.e. DARE cops made drugs sound cool) Program Evaluation Formative vs Summative A If B a program fails, it could be because the theory is wrong (more education does not reduce drug use) or because it was implemented incorrectly (i.e. DARE cops made drugs sound cool) Formative vs Summative Formative Main purpose: Evaluates how well pieces of logic model are fitting together. As program is implemented Often strengthen or improve the object being evaluated Video Clip: Office Space involves examining the delivery of the program or technology, the quality of its implementation, and the assessment of the organizational context, personnel, procedures, inputs, and so on. http://trochim.human.cornell.edu/kb/intreval.htm short term outcomes, those changes or benefits that are most closely associated with or caused by the programs outputs intermediate Long term outcomes or outcomes, those program impacts, follow changes that from the benefits result from an accrued though the application of intermediate outcomes. the short term Tip: List out your outcomes them put them in outcomes. chronological order.c Summative Main Purpose: Assessing whether Short, Intermediate, and LogTerm Outcomes are Met http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V7V-3WR4K4M-8&_user=735929&_handle=B-WA-A-W-AV-MsSAYZW-UUW-AAUDUAADDV-AAUVCEWCDV-YBACCWCBC-AVU&_fmt=full&_coverDate=06%2F30%2F1999&_rdoc=8&_orig=browse&_srch=%23toc%235852%231999%23999779998%23102970!&_cdi=5852&view=c&_acct=C000040778&_version=1&_urlVersion=0&_userid=735929&md5=6e6b6f92758532f01a88230d7ec293 80 Plot of Pretest and Posttest Means for a Possible Short, Median or Long-Term Outcome 1 step1 Step 2 A Step 4 step3 step1 Step 2 B Step 4 step3 Welch and Comer p. 9 Independent and Dependent Variables are implicit in Logic Models IV DV http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V7V-3WR4K4M-8&_user=735929&_handle=B-WA-A-W-AV-MsSAYZW-UUW-AAUDUAADDV-AAUVCEWCDV-YBACCWCBC-AVU&_fmt=full&_coverDate=06%2F30%2F1999&_rdoc=8&_orig=browse&_srch=%23toc%235852%231999%23999779998%23102970!&_cdi=5852&view=c&_acct=C000040778&_version=1&_urlVersion=0&_userid=73592 9&md5=6e6b6f92758532f01a88230d7ec29380 A Detailed Example of Design Notation Observations or measures (o) Treatments (or programs) (x) Groups (each is on a separate line) Assignment to a group R = random assignment N = nonequivalent groups C = assignment by cutoff Time (represented by moving from left to right) See Chapter 2 Welch and Comer for Research Design and p36 for Pro/Con Table Yuen and Terao 2003 :8 Applied Fact: The most widely used type of measurement procedure in program evaluation is the written survey completed by program participants. Tip: once you know when you will measure, decide what you will measure Yuen & Terao2003 Choosing the Scale of Measurement VARIABLE AGREE NO OPINION DISAGREE Applied Fact: The most widely used type of measurement procedure in program evaluation is the written survey completed by program participants. http://www.lafollette.wisc.edu/Courses/PA819/Categorical%20Data%20Analysis_Logistic%2 0Regression.ppt Yuen & Terao2003 Starting next class, We will watch two program overviews, for each we will: Watch video Practice Writing out notation Discuss issues of validity Practice the relevant stats approaches in lab form 1. 2. 3. 4. 2 stats approaches: Continuous Variables Categorical Variables Continuous Categorical You will use one or the other for each outcome (short, intermediate, long-term) Overview: To be prepared for next class you first need to know what types of outcome variables you have (just one each for short, intermediate, and long-term i.e. 3 total) Continuous Data Continuous Analysis Outcome Response Categorical Categorical versus Continuous Data Analysis http://www.lafollette.wisc.edu/Courses/PA819/Categorical%20Data%20Analysis_Logistic%2 0Regression.ppt Categorical Nominal Variables Variable: Kind of Beverage OR 1 2 3 1 2 3 Order anyway you please! http://www.lafollette.wisc.edu/Courses/PA819/Categorical%20Data%20Analysis_Logistic%2 0Regression.ppt Categorical Ordinal Variables Variable: Size of Beverage Small Medium Large http://www.lafollette.wisc.edu/Courses/PA819/Categorical%20Data%20Analysis_Logistic%2 0Regression.ppt Continuouspredict GPA Continuous Variable s ? associated with in. lb. Income Is weak 0 Weight Purchase associated with STRONG Height Income ? 1 Purchase http://www.lafollette.wisc.edu/Courses/PA819/Categorical%20Data%20Analysis_Logistic%2 0Regression.ppt ? Starting next class… For your final paper: Evaluation sections need to address 4 issues (and include design notation) Will you compare (measurement of objectives) by: I. Time? Pre-test, post-test Group? II. Comparison/control vs. treatment/intervention/experimental III. How have you considered internal validity? alternative explanations addressed IV. What statistical analysis will you use? Labs 5a and 5b Will watch two program overviews, for each will: 1. 2. 3. 4. Watch video Practice Writing out notation Discuss issues of validity Practice the relevant stats approaches in lab form Validity Does the instrument measure what it is meant to measure. External A reliable instrument yields the same results on repeated measures. Validity: the degree to which the conclusions in your study would hold for other persons in other places and at other times. (i.e. is it reasonable to believe that the program would work the same way in other communities?) Conclusion validity: the degree to which conclusions we reach about relationships in our data are reasonable, credible or believable. (i.e. don’t claim more than your data shows – if you note an increase in reading level, don’t assume this means an increase in grades or income) Trochim Text The Idea of Construct Validity “Does your program/ treatment accurately reflect/measure what you intended?” (i.e. you say more thorough baggage screening reduces terrorism, but are you actually measuring changes in screening and terrorism? Trochim Text Construct Validity When you measure what you term "self esteem" is that what you were really measuring? Content Validity: Whether a measuring device covers the full range of meanings or forms that would be included in a variable that is being measured. http://www.socialresearchmethods.net/kb/yinyangmap.htm For instance, self esteem can relate to many realms of people’s lives: work, school, family, etc. A content valid would include indicators (i.e. questions) that have to do with each of these settings. Source: Sullivan 2001 Methods of Social Research p.131-132 Internal Validity Internal Validity: observed changes can be attributed to your program or intervention (i.e., the cause) and not to other possible causes (sometimes described as "alternative explanations" for the outcome). (i.e.) there may be lots of reasons, other than your program, why test scores may improve or symptoms may reduce, What are they and how can you control for them? Baltimore Watch video Practice Writing out notation Discuss issues of validity Practice the relevant stats approaches in lab form 1. 2. 3. 4. Choose stats from chart Lab 5a compare means