PAGES:  [BACK to PROTEUS home page]    Back to CKHS100 MAIN PAGE     Outline     Research protocol

SPECIAL TOPICS:   Histograms/Central tendency    Sensitivity/Specificity   Rates   Critical review 

               Evaluation Research   Definitions    Designs  Sample size  Sample methods  Confounding 

ASSIGNMENT: A RESEARCH PROTOCOL

The final work for the course will be an opportunity to plan, assemble, and present a research PROTOCOL OR PROPOSAL, demonstrating good principles of research design.  

Select a topic, preferably from within your own area of expertise. The 'protocol' is essentially a complete design, detailed in all the steps and aspects of a piece of research without actually gathering the data.  You should demonstrate what the data might look like, and how it would be analyzed using "dummy" data.  This would be at the standard of an initial submission to a granting agency for funds to carry out the investigation. 

Your submission should follow a specific order and sequence of subject areas. This will still generally apply whatever the topic or design. (If a particular item is not applicable, then say so and give a short reason why it does not apply or is not needed)

The principal sections

1. STATEMENT OF THE PROBLEM that would have prompted you to do the research;  

2. BRIEF LITERATURE REVIEW: just enough to give an idea of the main 'schools' of thought. 

     Two-three paragraphs only. This requires an acceptable referencing  style (see example here);

3. YOUR RESEARCH QUESTION (RQ) AND ANY RELATED STUDY QUESTIONS

     This should follow naturally from your review of the literature. It should be a question!

4. MATERIALS AND METHODS (this will vary of course, but most protocols will include the following):

    * Target/study population - (what population are you interested in?)   

    * The design you are proposing:

            e.g.: Are you going to use a retrospective or prospective study? Cross-sectional or longitudinal?

            Intervention, RCT, experimental?  or Observational, case control, cohort?  Any qualitative components?

            Is this a new investigation or a replicate study to verify previous studies? Program/treatment evaluation?

            Patient compliance? 

    * Variables - type, and classification (Continuous or categorical? Independent or dependent?) 

    * Sources of bias and confounding that you can foresee in this type of study and how to overcome them

    * Data: proposed source; will you gather data yourself or use already-gathered data? Where from?

    * Sample methods for gathering data, (SRS systematic, etc., randomized from lists? volunteer lists?

                -Selection/rejection criteria? If controlling during selection, this could need stratification)

    * Sample size for confidence (survey) or a power analysis; how you would do these calculations?

               How many groups? Is stratification justified? Size of each group?

5.  ANALYSIS

    *  Any data transformation, tests of normality which might be needed?

    *  descriptive and/or analytic statistics to be used;

    *  Any controlling during analysis that might be necessary

    *  How your analysis would appear, (using 'dummy' data tables and examples of graphs, etc.);

6. CONCLUSIONS

    * What might you be able to conclude from the results and from the study?

    * How could the outcomes/results/findings address the original RQ (important!)?

    * Identify limitations and weaknesses in the methods used, including discussion

                  of any remaining potential sources of bias including confounders;

    * Recommended next steps in research on this topic.

 Written submission:  (Due date to be announced). Try to keep to 7-9 pages, including the lit-review, charts, tables, discussion, etc.  We are looking for a professional presentation that could be submitted to a research committee or funding agency. Correct referencing is essential in a proposal or other professional document.  You must attribute the source of the information, statement, etc.  The exception is your own material or observation, or  common knowledge, such as "obesity is more commonly seen today".   Here is a style guide, or you can use another (e.g. APA).  

Oral presentation:  In previous cohorts, the everyone was asked to prepare a very brief (6-7 min) presentation of their protocol during the final session.  We are not sure this can be done this time with the rearranged schedule.   More details to be announced.       

Examples of topics from previous groups  

The best approach is SIMPLE and NARROW. Don't attempt a wider topic.

   * When encouraged to begin losing weight, do male and female patients demonstrate similar compliance / completion / success rates,  and is there an effect due to different age groups, or initial BMI groups?

   * How does family behaviour influence young people re smoking?  What are the motivations to smoke/not smoke when parents currently smoke? when parents have given up smoking or are non-smokers?  Does identification of smoking related illness in friend or family member influence behaviour?

   * (Using a questionnaire survey) ... what proportion of pts 'know' about correct procedures with prescriptions, and what proportion actually PRACTICE the correct actions?   (e.g.  avoiding keeping old/expired prescriptions around the house?, always taking a complete course of antibiotics?, giving remains of old prescriptions to other family or friends?, self medicating with family/friends' old antibiotics?, placing pills into bottles without labels or wrong labels?, ... etc. etc.)  With males/females, old/young comparisons.

   * (using a questionnaire survey) ... what proportion of pts actually have some physical exercise daily? or how many hrs/day sitting?  [and males/females, old/young comparisons]

   * (Using a questionnaire survey) ... what proportion of pts know what the signs are of a family member having a stroke? - an MI?  - choking?  etc., and would know what to do?   [and make male/female, old/young,  trained/untrained comparisons]