COURSE DESCRIPTION:
Targeted to the needs of graduate students, this course compliments and expands on existing mathematical-based instruction with practical, “plain English” explanations in order to provide the skill requisite to applying and interpreting statistical concepts appropriately. Actual data sets (provided by neuroscience faculty, open data sources and the students) and hands-on data visualization and analyses in each session provide real-world examples of the typical and unique challenges faced in experimental science. Assignments are practical elements of a qualifying exam, grant proposal or research paper.
COURSE OBJECTIVES:
1) Practical experience, through the use of actual data-sets (detailed below), in choosing appropriate data analysis tools and statistical models for typical data encountered in neuroscience studies.
2) Choosing the Correct Statistical Tests: Mastery of the implications, pros and cons, assumptions and limitations of various statistical models. Effective and transparent data visualization and illustration.
3) Application of statistical principles to pre-registration and experimental design (including power analysis).
4) Learning to correctly report statistical data – learning to interpret and understand statistical data reporting and identify errors and false assumptions in published and presented data.
5) Avoiding common pitfalls.
6) Application of these basic principles to complicated data sets for the greatest transparency and rigor.
PREREQUISITES:
None. Undergraduate statistics recommended.
REQUIRED MATERIALS:
- Computer (either platform) – contact the course director if this is an issue.
- JMP (SAS) cost $29.95.
SUITABLE FOR 1ST YEAR STUDENTS:
Yes. Priority enrollment is given to grad students, but postdocs and other are welcome if the max enrollment has not been reached.
STUDENT ASSESSMENT:
Weekly HW assignments – 30% of grade
1 data management project, 1 data analysis and visualization project, 1 submission of dataset, 2 take home exams (in the form of mock qualifying exams, mock grant proposals or mock publication) 70% of grade (one mid, one final)
Feedback given for each assignment and in class. Problem sets with feedback in class.
Other formative assessments include – daily or weekly reflections on Canvas, short summaries of the main points covered in that week. Assignments may include webinar type instructional videos for how to use the software (although in class instruction is provided).
Attendance and Participation:
No more than one unexcused absence per session. All absences (excused or otherwise) must be “made-up” by completing the requisite work. No more than 3 excused absences per session.
The data sets and resulting analysis should have been completed during each class session, with appropriate guidance to a reasonable standard. Students will be required to upload these to Canvas
Objective Assessments:
- Summative assessments = 2 exams, 1 data management and simulation spreadsheet, 1 data analysis and visualization project.
- Formative assessments - weekly HW and in class exercises in software use, data analysis and data management – includes Peer reviews, presentation, and critical analysis of published papers, and submission of sample data in your field.
Subjective assessments:
Will be assessed before and after the class.
- Familiarity with, and confidence in judging and evaluating the suitable models available for statistical analysis.
- Familiarity with and confidence in evaluating statistical data reported in seminal and relevant publications.
- Completes the worksheets, analysis in class correctly and demonstrates sufficient familiarity with software features
- Reflections will be requested weekly.
It should be noted here that despite the fact that many students have a decent rote understanding of the mathematical formulae and assumptions underlying basic statistics, they do not recognize these in the real world and express a profound (and accurate) lack of confidence about the application of these basic principles in their daily work.
CREDIT HOURS: 2.5
