research methods lecture 1 objectives today

27
Research Methods Lecture 1 Objectives Today • Orientation to class • Orientation to Professional Expectations • First lecture on content • First lab • First homework (due today)

Upload: kyla-guzman

Post on 31-Dec-2015

22 views

Category:

Documents


0 download

DESCRIPTION

Research Methods Lecture 1 Objectives Today. Orientation to class Orientation to Professional Expectations First lecture on content First lab First homework (due today). PROFESSIONALISM IN PHYSICAL THERAPY: CORE VALUES. 1. Accountability 2. Altruism 3. Compassion/Caring 4. Excellence - PowerPoint PPT Presentation

TRANSCRIPT

Research Methods Lecture 1Objectives Today

• Orientation to class• Orientation to Professional Expectations• First lecture on content• First lab• First homework (due today)

PROFESSIONALISM IN PHYSICAL THERAPY: CORE VALUES

• 1. Accountability• 2. Altruism• 3. Compassion/Caring• 4. Excellence• 5. Social Responsibility• 6. Integrity• 7. Professional Duty

First RM Lecture

Is the data relevant?

What is wrong with this question?

Subject verb agreement

First RM Lecture• Test/Measure – used to obtain a measurement

• An attribute – a variable, a characteristic, a quality

• An examination – a test or tests

• An evaluation – a judgment

• Operational definition – provides a common framework

Clinical Research• Quest to be Evidence-Based Profession

• Our Practice (every patient we treat) should be– Backed up by research/data– Using the current best evidence

• Why?

Clinical Research

Structured process

Empirical and critical= able to observe it

Framed by our societal/professional paradigmsOutcomes importantHealth and disability models are frame of referencePractice should be Evidence-Based

Research is Careful, Logical, and Systematic

Clinical Research

• Research is Careful, Logical, and Systematic

• It does not prove• It can lend support (for or against)

• Research is repeatable– Must be to be reliable and hence valid

Measuring Outcomes Did the patient live? Did impairments lessen? Did functional limitations lessen? Did health improve? Did we save money? Did we save time?

Health Related Quality of Life (HRQOL)SF-36WOMACODI

Models of Health and Disability

Nagi modelPathology, Impairment, Functional Limitation, Disability

International Classification of Function “Ability” slant to the Nagi model

Capacity vs Performance

Provide a framework for relevant outcome measures

Figure 1.1 in P & W good summary models

ICF Model

Evidence-Based Practice

• Sackett a giant in this field• “integration of best research evidence with our

clinical expertise and our patient’s unique values and circumstances”

• Ask an answerable question• PICO one way of framing a question

• Then look to the literature

One Way to Ask a Question

• Population of interest

• Intervention you are considering

• Control/Comparison Treatment

• Outcome that you will measure

Sources of Knowledge• Tradition: least powerful*• Authority• Trial and Error• Logic– Deductive– Inductive

• Scientific Method Assumptions: most powerful*– Nature is orderly– Events not random– Limitations in our profession?

Research A continuum (Fig 1.5 P&W)

Case Study Randomized Controlled TrialsEverything in between

Research process Identify the question/topic (general) “PICO”Design the Study

Search the literature Define the topic (specific) State a general question Operationally defined hypothesis

Methods (collect and reduce data)AnalyzeCommunicate

Measurement• Continuously– How tall are you?

• Discretely– What is my heart rate? 60 ¾ bpm vs 60 bpm– Are you male or female? XX XY YY, more than 2

• Dependent of our ability to define• Measurement often indirect= HR measured in pulse or BP is in cuff

• Precision: number of decimal points vs. Accuracy:• Amount and units (22 inches) need to have units

Nominal Dataaka Categorical Data

• “3 vs 4 categories”• Finite number of classifications= can code with a label

for ease, but has no meaning. • Categories• Mutually exclusive= cant be both male or female,

one result.• Exhaustive• Numbers or symbols may be assigned

– Have no inherent value– Can be changed without affecting the data

• Counting is the only math allowed! = no averages taken

Ordinal Data

• Rank ordered, 1st step, 2nd step, etc.• Those of the same rank not required to be equal• Ranks are not equally spaced• May or may not have true zero point• Averaging is allowed– But the value doesn’t reflect a true quantity

• Most common ordinal scale is Likert scale“1 = Very satisfied, 2 = satisfied, 3 = neutral….”

*not a true quanity

Continuous Data“Scale” Data in PASW1 scale covers 2 scales

• Interval and ratio data are here– Difference is presence of true zero, interval has no 0– Like the temperature scale 0, 32, 273

• Can be broken into smaller pieces –continuously

Interval Scale

• Rank ordered• Ranks are equally spaced• No true zero point• Transformation cannot be done directly• Transformation can be done while retaining

relative positions• Mathematical operations should not include

intervals because there is no true zero)

Ratio Scale

• Rank ordered

• Ranks are equally spaced

• There is a true zero point

• All mathematical operations permissible

The Research Question More important than ability as statistician Requires refinement to answerable question

TopicResearch ProblemResearch Question

Research Question shaped by:Clinical experience (no substitute for this)Feasibility and importancePopulation to be sampledRationaleVariables to consider

The Variables• Independent (IV in this course)– The predictor = usually the treatment or time

– Can have more than one– Always at least two levels for each independent variable

• Dependent (DV in this course)– The outcome– Can have more than one DV= multiple outcomes, ROM, Strength, pain

– No levels, EVER, EVER, EVER (not even 1)

In class practice=both for most

• Strength=either• Nerve Conduction Velocity=dependent• Age=independent• Employment status=both; predict scores on

disability index

The Hypotheses Research hypotheses

True expectation What you think is the answer to your question There will be a difference from intervention, but can be a dull sometimes

Statistical (null) No difference/no relationship between IV and DV, when we want to show that

there is no difference like drug vs exercise

Directional (difference will be in a specific direction) easier to find a difference in A will be bigger than B

Non-directional (only that there will be a difference) is there difference in either smaller or bigger

Simple (One IV (infinite levels) and One DV (infinite outcomes, no levels!!)) Complex (more than one IV and or DV)

The Literature• Review it every step of the way (Fig 7.1 P & W)

• Differentiate your sources– Primary sources• Peer reviewed journal articles• Oral presentations from the researcher• Interviews• Personal experience

– Secondary sources• Textbooks (good source of citations of primary sources)• Review articles

• Measurements• NIOR