Identifying a Research Design Type:
You can choose any research design type (Design Declaration) in Chapters 15 through 18 in the Research Design in the Social Sciences BCH book. However, different research design types will have different potential rubric points – i.e. the maximum grade for some research design types will be higher than for others. The logic here is that some research designs are more difficult to implement than others, and should be evaluated as such, but even being able to implement the most basic research design successfully should result in passing completion of the assignment. The designs (by BCH Declaration Number) are placed into categories and potential rubric points are given in the table below:
Category
Designs
Maximum Points
A
All Chapter 18 declarations except 18.1 & 18.2; All Chapter 16 declarations except 16.1; 17.2; 15.5 (which requires 15.4)
100
B
All Chapter 17 declarations except 17.2; 15.4; 15.6
90
C
15.1; 15.2; 15.3
80
RUBRIC
Originality (Maximum 10 points of all categories)
You can choose to either:
Replicate the design declaration verbatim in code and narrative (0 points)
Modify the design (in narrative and/or code). The number of points you receive (up to 10) will be determined by how substantially you modify the design that is in the book.
Using the tutorial/practicum modifications will receive partial credit.
Research Design Narrative ~ 500-1000 words (approximate). (Maximum 45 points category A, 40 points category B, 35 points category C).
Your research design narrative should describe, in words, all of the applicable elements of your research design including your model, inquiry, data strategy and answer strategy (Not all research design will have all elements – you will be evaluated on the quality of narrative for those elements present in your chosen research design). Your Model should describe, as applicable, your units, endogenous and exogenous variables, and (potential) outcomes. You are encouraged to include a Directed Acyclic Graph (DAG) which shows relationships between variables that your theoretically justify in a written narrative. The Inquiry should describe your research question and estimand, and note if the target of the inquiry is descriptive or causal. Your Data Strategy should describe any applicable sampling, treatment or measurement strategies. Your Answer Strategy should describe your estimation or case study approach and why that answer strategy is appropriate for your inquiry.
DeclareDesign in R, coding and diagnosis (Maximum 45 points category A, 40 points category B, 35 points category C).
The second stage of your research design is to translate your narrative design into code using DeclareDesign “declare” functions and diagnosis that design using diagnose_design. Materials submitted for this portion of the assignment should include a R script file of your design in code, a R history file which demonstrates your code runs successfully, and one or more graphical representations that illustrate one or more diagnosands from your design (estimates, estimands, power, rmse, se, etc.). These could be as simple as a histogram of your estimates from a number of simulated diagnoses, or it could be as complex as plotting multiple power curves or ACMEs over a range of different model parameters. You should also include a short (~100-200) word addendum section to your Research Design Narrative titled “model performance” that reflects on some aspect of your model diagnosands.
Submission Checklist
Research design narrative, including discussion of model diagnosands
R Script File
R History File
1 or more graphical representation
