Classes

clp_course_ladder

Below are links to syllabi for classes that I teach.  You may contact me at marsiske@phhp.ufl.edu . At the bottom of this page, I also list some courses/programs in other departments that might be of interest to students.

Course Title When
offered
Most
recent Syllabus
Pre-requisite
 CLP 6527 stats_image1 Measurement, Design and Statistics I every Fall syllabus Admission to graduate study in  Clinical and Health Psychology, Psychology, Rehabilitation Sciences,Health Services Research, Management and Policy, or Speech, Language and Hearing Sciences. All others must petition (see below). To opt out of CLP 6527 (for those for whom it is required), click this link.
 CLP 6528 Blackboard with mathematics sketches - vector illustration Measurement, Design and Statistics II every Spring syllabus Completion of CLP 6527 with a grade better than C+. All others must petition (see below).
CLP 6529 stats_image4 Applied Multivariate Methods in Psychology every Fall syllabus Completion of CLP 6528 with a grade better than C+. All others must petition (see below).  EPI and MPH students who have taken taken PHC 6052 and PHC 6053 will have a strong probability of successfully petitioning.
 CLP 7525 stats_image3 Best Methods for the Analysis of Change every Spring syllabus Completion of CLP 6529 with a grade better than C+. All others must petition (see below).
 CLP 7934 np_image Neuropsychology of Aging directed reading
class; every
semester)
syllabus None.  Prior coursework in cognitive psychology and/or neuropsychology is strongly recommended

 

course_overview

Petition for CLP 6527 and CLP 6528

  • Send an email that answers the following questions along with this petition form.
    • – describe your past statistics training, both undergraduate and especially graduate
    • – describe the currency of your training; specifically, the longer it has been since your previous coursework, the more we will want to know about how you have practiced these skills and kept them current
    • – explain why other available courses at UF cannot satisfy your training needs
    • – discuss, specifically, your background and comfort with the SPSS program
    • – describe your level of comfort with mathematics, statistical formulae, and theoretical statistics; if you are not comfortable with these areas, describe how you will manage that discomfort so it is not an impediment to your learning

To Opt Out of CLP 6527/6528

  • Send an email that answers the following questions, after reviewing the syllabi for CLP 6527 and CLP 6528 above:
    • – Whatare the perceived areas of overlap? Why do you think this course would not be a useful use of your time?
    • – If  your argument is ‘I think I have taken this before’, I’d also like you to address WHEN you took the class (how long ago).  Assuming you didn’t finish the course last month, I’d like you to explain how you kept your methodological and statistical skills FRESH and ACTIVE in the period of time between when you last took the course and  now.  Also, my pedagogical bent says “overlearning is good”.  Especially in statistics and methodology, where students don’t always claim to have the highest levels of comfort, explain why you don’t think hearing the material again (especially from a different perspective, in different words, from a different instructor) wouldn’t be good for you as a scientist.
    • – Past experience has shown us that there are usually two reasons for students to want to opt out.  The first is because they had the material before (as above), but the second is because—all things being equal—they really don’t enjoy the stats stuff.  The problem is that, as the research progresses, these students often end up being behind their classmates because their lack of enjoyment/motivation, in the past, turned out to have harmed  their retention and learning.  Convince us why a lack of motivation/lack of enjoyment is not the chief driver of your opt-out request.
    • – Tell me whether your previous course emphasized SPSS or SAS.   Here at UF, SPSS is the more commonly available program, so I have chosen to teach the class using mostly SPSS.  If you previously trained in SAS, explain how you think (outside of taking the class) you’ll get the appropriate level of comfort and familiarity with SPSS.  If you’re comfortable with SPSS, just reassure me about that.
    • – How would you decompose two- and three-way interactions in an analysis of variance? (this is a good place to talk about simple effects decomposition or related planned/post hoc strategies)
    • – How do you assess whether there is a problem with regression outliers in a multiple regression? Describe the quantitative and qualitative/visual tools you would use, and what decision criteria you would apply.
    • – How do you determine whether a variable distribution is appropriate for any general linear model procedure? If you discover a problem, what transformations or non-parametric approaches might fix it?
    • – What quantitative tools do you use to determine if there is a problem of predictor multicollinearity, and how do you fix it?
    • – How can  estimate an interaction between two continuous predictors in a regression model, while avoiding a serious multicollinearity problem? (Your answer should NOT propose categorizing a continuous variable; this is a good place to follow the recommendations of Aiken & West, or discuss residual centering).
    • – Using a 2×2 table, how can you compute and interpret a   simple odds ratio? How does this relate to a logistic regression?
    • – Using a 2×2 table, comparing a diagnostic test and a gold standard diagnosis, how can you compute sensitivity and specificity? How do these relate to a Receiver Operating Characteristic?
    • – What might be the problem with test-retest stability as an index of an instrument’s reliability? What might be better approaches for assessing reliability?
  • Please note, I will simply make a recommendation to your area head or training director.  They will make the ultimate decision.


Petition for CLP 6529

  • Send an email that answers the following questions along with this petition form
    • – describe your past statistics training, both undergraduate and especially graduate
    • – describe the currency of your training; specifically, the longer it has been since your previous coursework, the more we will want to know about how you have practiced these skills and kept them current
    • – explain why other available courses at UF cannot satisfy your training needs
    • – discuss, specifically, your background and comfort with the SPSS and AMOS programs
    • – describe your level of comfort with mathematics, statistical formulae, and theoretical statistics; if you are not comfortable with these areas, describe how you will manage that discomfort so it is not an impediment to your learning
    • – How would you decompose two- and three-way interactions in an analysis of variance? (this is a good place to talk about simple effects decomposition or related planned/post hoc strategies)
    • – How do you assess whether there is a problem with regression outliers in a multiple regression? Describe the quantitative and qualitative/visual tools you would use, and what decision criteria you would apply.
    • – How do you determine whether a variable distribution is appropriate for any general linear model procedure? If you discover a problem, what transformations or non-parametric approaches might fix it?
    • – What quantitative tools do you use to determine if there is a problem of predictor multicollinearity, and how do you fix it?
    • – How can  estimate an interaction between two continuous predictors in a regression model, while avoiding a serious multicollinearity problem? (Your answer should NOT propose categorizing a continuous variable; this is a good place to follow the recommendations of Aiken & West, or discuss residual centering).
    • – Using a 2×2 table, how can you compute and interpret a simple odds ratio? How does this relate to a logistic regression?
    • – Using a 2×2 table, comparing a diagnostic test and a gold standard diagnosis, how can you compute sensitivity and specificity? How do these relate to a Receiver Operating Characteristic?
    • – What might be the problem with test-retest stability as an index of an instrument’s reliability? What might be better approaches for assessing reliability?

Petition for CLP 7525

  • Send an email that  answers the following questions along with this petition form
    • – describe your past statistics training, both undergraduate and especially graduate
    • – describe the currency of your training; specifically, the longer it has been since your previous coursework, the more we will want to know about how you have practiced these skills and kept them current
    • – explain why other available courses at UF cannot satisfy your training needs
    • – discuss, specifically, your background and comfort with the SPSS and AMOS programs
    • – describe your level of comfort with mathematics, statistical formulae, and theoretical statistics; if you are not comfortable with these areas, describe how you will manage that discomfort so it is not  an impediment to your learning
    •  – What are the ways in which a MANOVA differs from  multiple ANOVAs in its handling of a set of dependent variables?
    • – How might a discriminant function analysis be used as a followup or post-hoc procedure to better understand the results of a MANOVA?
    • – In mixed effects models, if an IV has both a fixed and a random effect, what does this mean? What different information do the two types of effects provide?
    • – In mixed effects models, what is the difference between a Level 2 vs. Level 1 predictor? What different information do the two levels of predictors provide?
    • – What is Kaiser’s rule for determining a factor structure in exploratory factor analysis?
    • – When might correlated uniqueses/residuals be allowed in a confirmatory factor model?
    • – What is the concept of factorial invariance, and how is it  assessed?

Programs/courses in other departments