Download presentation
Presentation is loading. Please wait.
Published byTrevor Maximillian Ford Modified over 9 years ago
2
How to infer causation: 8 strategies? How to put them together? S519
3
Adding value to descriptive data to make our evaluation explicit Our goal Using quantitative value to evaluate the quality or value of the evaluand in a particular context. Build up our conclusions based on a level of certainty What are values: „good“, „valuable“, „worthwhile“ S519
4
Adding „values“ to descriptive data collected about Process, outcomes, costs, comparisons, exportabilities; or Situated dimensions or components Weighting all the strengthens or weaknesses of these values to draw overall conclusion about the evaluand. How Importance weighting Merit determination synthesis methodology S519
5
Before we go for methodology, we have to answer the question: Whether our data are subjective S519
6
Subjective 1: Inappropriate application of personal or cultural preferences/biases: arbitrary, idiosyncratic, unreliable, highly personal (i.e., based purely on personal preference, cultural biases, gender biases) Subjective 2: assessment or interpretation by a person, rather than guidelines Using well-founded expert judgments Robust evidence Subjective 3: about a person‘s inner life or experiences (e.g., headaches, fears, beliefs, emotions, stress) Usually not independently verifiable S519
7
We provide our conclusion based on certainty in the relevant decision-making context Keep the whole evaluation well documented and justified All evaluations, especially high-stakes ones, should be meta-evaluated (i.e., evaluation itself should be evaluated) S519
8
Importance determination is the process of assigning labels to dimensions or components to indicate their importance. Importance weighting Prioritize improvements Identify whether identified strengths or weakness are serious or minor Work out whether an evaluand with mixed results is doing fairly well, quite poorly, or somewhere in between. S519
9
Different evaluations Dimensional evaluation Looking at multiple dimensions of merit that pertain to the evaluand as a whole rather than separately to its parts. Component evaluation Looking at each of the evaluand‘s components (or parts) separately and then synthesizing these findings to draw conclusion about the evaluand as a whole. Each component can be evaluated on several dimensions that pertain to this component only rather than to the evaluand as a whole. Holistic evaluation Looking evaluation as a whole without division into dimensions or components S519
10
Component analysis Evaluating policies, programs, or interventions that have several quite distinct parts An international program consisting of projects implemented in different locations (e.g. „WIC“ in IU) A government policy includes multiple policy measurements (e.g. Juvenile delinquency) An organizational transformation includes several distinct interventions (e.g. Career support) S519
11
Dimensional evaluation Entities whose quality or value is experienced by consumers on multiple dimensions that pertain to the evaluand as a whole Product evaluation (i.e. Car evaluation) S519
12
Holistic evaluation Unusual in the evaluation of programs, policies and other large complex evaluands. More common in personnel, product and service evaluation (expertise-oriented evaluation) Judging the overall quality of a sample of writing Grading essays Classroom teaching Athletic performance cosmetics S519
15
Weak performance on minor criteria (e.g. dimensions, components) may be no big deal, But weak performance on important criteria can be very serious issues. S519
16
1. having stakeholders or consumers „vote“ on importance Commonly used in both participatory and nonparticipatory evaluations Collecting opinions from everybody Assumptions Each person is well informed Stakeholder‘s belief what (s)he chooses is important Stakeholder‘s important should be treated equally Pros and cons? S519
17
2. Drawing on the knowledge of selected stakeholders Using selected stakeholder input to guide the assignment of importance weightings Collecting opinions from selected experts Setting up the Bars A bar is a defined minimum level of criterion performance below which the evaluand is considered completely unacceptable, regardless of performance on other criteria. S519
18
2. Drawing on the knowledge of selected stakeholders Assumptions: The stakeholders should be sufficiently well informed to provide valuable relevant information The combination of stakeholder input will provide sufficient certainty about importance for the given decision-making context Pros and cons? S519
19
3. Using evidence from the literature Literature review Evaluations of similar evaluations in similar contexts Research documenting the key drivers (or strongest predicators) of success or failure with this type of evaluand. Assumptions The volume and quality of the available research is sufficient to judge the importance The context of other research is sufficiently similar to yours and therefore that the findings can be reasonably applied to your setting Pros and cons? S519
20
4. Using specialist judgment When you have tight timeline, no time for gathering stakeholders and looking for literature Identify one or two (or two or more) well-known specialists in the domain Better be supplemented with other evidence Pros and cons? S519
21
5. Using evidence from the needs and values assessments Determining the importance of criteria (dimensions) Any frequently mentioned characteristics? Looking for poor-performing evaluators that cause serious problem Looking for top-notch evaluators that have dramatic impacts on success S519
22
5. Using evidence from the needs and values assessments Determining the importance of components Severity of dysfunction addressed (primary consideration) Scarcity of alternatives: no other options for addressing the need. Intent to use alternatives: if the evaluand component in question did not exist. Rubrics to measure (Table7.4, 7.5 and 7.6, 7.7, 7.8 (combined)) Prons and cons? S519
23
6. Using program theory and evidence of causal linkages When criteria or components are linked to needs through a complex logic chain. Such as „soft“ skills or attributes (e.g., inspirational leadership, self-esteem, stress management, a kind of instrumental needs) More upstream variables (see Exhibit 7.5) How to estimate the strengths of the links Interview Analyze your previous data ... S519
24
Always think whether they are applicable Choose mulitple of them S519
25
Table 7.10 (class dissusion) Form a group Discuss which strategies you will choose to determine the importance for the “student services in the school health program” (see Table 7.8) Discuss which strategies you will choose to determine the importance for your group project S519
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.