Sampling and Data Collection Research Methodologies in Allied Health Peg Bottjen, MPA, MT(ASCP)SC
Research Process Planning Hypothesis or Aims Research Design Data collection The research process involves: Planning for the project Forming a hypotheses Selecting a research design Collecting the data Organizing the data and presenting the data Analyzing the data with statistical methods Interpreting the data and forming conclusions At this point depending on the findings, revising or adding to the research project may be necessary and that goes back to the planning stage. You will be going through this process as you work in your groups to prepare a research proposal. Organization and Presentation of Data Data Analysis Interpretation and Conclusions
Objectives Upon completion of this unit the student will be able to: • Outline advantages and disadvantages of three sampling methods • Operationally define the variables in a study • Identify limitations in research studies • Choose the method of data collection that best facilitates proving or disproving a hypotheses or best answers a research question • Explain how a test can be valid yet not reliable.
Boundaries of the Study Define the terms Dictionary definitions ex. BMI – wt (kg) / ht2 (m) Reference to a definition ex. NHANES II study Operational definitions ex. obesity >95th percentile
Operational Definition The operational definition of a variable is a statement of how the researcher in a particular study chooses to measure the variable in question.
More Boundaries of the Study Assumptions Scope Limitations Subjects
Sampling Target Population Representative Sample X X X X X
Representative Sampling Methods Randomized – roll of dice Area - location Systematic – choose from list (every 5th) Volunteer – ask for participants Convenience – what is available
Subjects Inclusion Exclusion
Sample Size It depends! Procedures used to collect data Statistical level of significance Statistical power Effect size
External Validity The extent to which the results of an investigation can be generalized to other samples or situations. Polgar & Thomas-4th Ed. 2001
Internal Validity The intervention was responsible for the differences or lack of differences observed.
Data Collection Methods Questionnaire Interview Observation Measurement / instrumentation
Questionnaire Define information needed Draft questions Do pilot Redraft Administer questionnaire
Questionnaire Open vs Closed questions Scaled questions or statements Likert scale Strongly disagree, disagree, no opinion, agree, strongly agree Semantic differential Draw an x at point that best describes your feeling Powerless Powerful X
Interview Structured Unstructured
Observation Who is observer Setting of observation Use of instrumentation
Observation Complete participant Participant as observer Observer as participant Complete observer
Measurement / Instrumentation Subjective measurements Ratings or judgments by humans of quantities or qualities Objective measurements Measurement of physical quantities or qualities by equipment
Measurement / Instrumentation Objective measurement Data collected by instrumentation Eg. Blood glucose, MR/CT scans, radiographic image, nuclear medicine scan, activated clotting time (ACT), nutrition screening instrument, stained slide read by expert Data collected by standardized tests Eg. SAT, ACT, GRE, CAT, Myers-Briggs, IQ test, patient quality of life (SF-36)
Desirable Properties of Measurement Tools Reliability Validity Applicability Practicality
Reliability - Precision The test or assessment measures the same thing every time and results in the same answer every time. Test-retest – correlates 1st to 2nd Inter-observer – between different observers Internal consistency – between items measured
Validity - Accuracy The test or assessment measures what it claims to measure. Sensitivity - % true positive Specificity - % true negative Test Result Disease Present Disease Absent + True + False + - False - True -
Validity of Mammography Mammography Result Known Breast Cancer + 169 ( true +) - 49 (false -)
Validity Predictive value Ability of a measurement to predict the value of it or another test in the future (prognosis) What is predictive value of an ACT score to your grade point average in college?
Summary Boundaries of the study Sampling Data Collection Methods Reliability/ Validity
Critical Evaluation of Research Who comprised the sample? What was the inclusion criteria? What was the exclusion criteria? Was the sample biased? How is the independent variable defined? How is the dependent variable defined? How were validity issues addressed? How were reliability issues addressed?
Group Work Questions 9 – 16 in the Research Proposal Notebook are designed to lead you through the process of the literature review, defining the research question, and selecting a research design. Answer each question in your group. Submit your results as described under Assignments.