
Components of Informed Consent
 1. Purpose of the study
 2. Procedures
 3. Potential Benefits
 4. Potential Risks
 5. The fact that they may withdraw at any time without penalty

Importance of Informed Consent
the key to promoting ethical values in research
 ASHA Principle I of Ethics
 Rules of Ethics N: Individuals shall use persons in research or as subjects of teaching demonstrations only with their informed consent

Institutional Review Boards (IRB)
 independent bodies that review research for the protection of human participants
 approve plans that eliminate/reduce harm and offer some benefit to the participant
 evaluate risks to participants, assess safeguards, and recommend modifications when necessary

Levels of Review for IRB
 Full review: the standard
 Expedited review: faster process
 Exemption: the study doesn't require it because there are no subjects (ex. metaanalysis)

The Belmont Report
 developed framework for ethics code
 institutions must be clear about routine and research services purposes
 vulnerable populations must be protected
 research must demonstrate respect for persons, beneficence, and justice

3 Critical Elements of the Belmont Report
Respect for persons: recognize people have value and choice (informed consent)
Beneficence: do no harm, maximize benefits to the client, try to make benefits outweigh the risks
Justice: treat all people equally; subject selection formed to the study and fair

Components of IRB Application
Contact Dr. Adibi
 approval form
 abstract
 informed consent agreement
 support letter(s)
 survey instrument, treatment protocols, test instrument, test/treatment materials

Abstract/Summary of IRB
Intro (rationale, statement of purpose, hypothesis)
Method (design, setting, participants, instruments, procedures for recruiting subjects, data collection procedures)

Evidence Based Practice (EBP)
 process that aims to provide clients and practitioners with info needed to choose the best procedure/treatment for the client
 current high quality research, clinical experience, and client preference is used to make treatment and diagnostic decisions

Components to Ensure EBP
 clinical decisions based on most up to date scientific and clinical evidence
 clinician level of experience
 involve the client in the decision making process

Steps of EBP
 1. ID the clinical problem
 2. gather info about the problem from research studies
 3. ensure adequate knowledge to read and analyze the studies
 4. summarize info to use in your practice
 5. define expected outcomes for clients and the families
 6. provide EDU and training to implement the suggested change in practice
 7. eval the practice; change/modify if necessary

2 Major Sources of Info for EBP
 raw evidence
 prefiltered evidence

Raw Evidence
info that has not been subjected to expert review

Prefiltered Evidence
info that has been reviewed by experts

History of EBP
 governed by medical EBP
 EBP began in audiology first (1995)
 2005 ASHA convention dedicated to formally make SLP EBP
 ASHA code of ethics for EBP made in 2003

Ethics in EBP
 must include client's best interests
 must favor the proven intervention
 based on clinically relevant evidence in collaboration with the client
 considers costs, benefits, resources, and options

Model For EBP
 1. Formulate answerable question about prevention, diagnosis, prognosis, or therapy
 2. Locate best evidence to answer the question
 3. Eval the evidence based on validity, impact, and utility
 4. Combine eval with clinical experience and the client's unique circumstances
 5. implement and eval effectiveness and efficacy

PICO Questions
Patient/Population Intervention Comparison and Outcome questions
 population should include age, disorder, severity, time since onset
 intervention of interest stated in a specific way
 comparison of other possible interventions for the population in question
 what are the expected outcomes?

Things to Remember about EBP
 Every case is different
 Not solely a matter of science and evidence; recognize specific client needs
 reading textbooks/journals and attending conferences is not enough for EBP
 EBP can be effective without randomized clinical trials

Measurement
systematically assign numbers to object, persons, or events according to rules determined by the research design to determine the degree of difference
 determine the extent of IV on DV
 allows us to examine sim/diff between measured events
 forms the basis of statistical analysis

Data
 form (usually numerical) in which measurements are collected and stored
 dictate the types of statistical analyses that can be applied

Parameter
a number describing a population characteristic
ex) average weight of women  based on gen pop, not sample

A Statistic
number describing a sample characteristic

Two Types of Statistics
 Descriptive Statistics
 Inferential Statistics

Descriptive Statistics
 involves tabulating, depicting, and describing data
 a specific feature or characteristic of a set of data is measured
 data summary used to organize data (visual description of the results)
 used in qualitative and quasi experimental studies
 may use simple count data

Simple Count Data
 number of occurrences of the behavior
 used in small samples

Inferential Statistics
tests that allow us to estimate/predict characteristics of a pop from knowledge of characteristic from the sample
(make inferences to the gen pop)

4 Scales of Measurement (Stevens Taxonomy)
applied to data dependent on its properties
 Level 1: Nominal Level
 Level 2: Ordinal Level
 Level 3: Interval Level
 Level 4: Ratio Level

Nominal Level
 Level 1 of Stevens Taxonomy
 naming level
 has no quantitative properties
 allows us to classify groups, categories, behaviors, and events
 assumed that groups don't overlap
ex) male and female; may be numbered (male=1 and female=2) as a descriptor, but the number itself has no value

Ordinal Level
 Level 2 of Stevens Taxonomy
 to rank/order the levels of the variable being studied
 imprecisely measured data
 ranked highest to lowest, greatest to least, etc.
 not suitable for statistical analysis
 ex) the likert scale

Interval Level
 Level 3 of Stevens Taxonomy
 used to analyze inferential statistics
 meaningful difference between the numbers on the scale
 intervals are equal in size
 distance between intervals is known and fairly consistent
 NO ABSOLUTE ZERO (zero does not mean the absence of something)
 may have a relative zero
 ex) decibels  zero decibels isn't the absence of sound, just the lowest sound a human can hear

Ratio Level
 Level 4 of Stevens Taxonomy
 highest level of measurement
 used in inferential statistics
 same characteristics as interval level PLUS the presence of an absolute zero
 can compare points along a scale in absolute terms
 continuous variables that have all the math properties

Nominal Variables
 used to name things
 no numerical value

Numbers on an athlete's jersey represent what type of variable?
Nominal Variable

Measures of Central Tendency
 mean (average)
 median (middle)
 mode (occurs most often)
can always be applied to data

What does a negative distribution do when compared to a normal distribution?
underestimates performance of the participants

What does a positive distribution do when compared to a normal distribution?
overestimates performance of the participants

When is the best time to use the mean when describing data?
when the distribution is normal

Nonparametric Statistical Analysis
 does not need normal distribution
 does not assume that data reps a normal distribution
 used on nominal and ordinal data

Parametric Statistics
 analyze groups with normal distribution
 used for interval and ratio data

Standard Scores
 raw scores that are converted to standard deviation units
 also known s zscores

Raw Score
 initial score on a test
 has no real value
 must be changed to converted (standard) score

Levels of Power Applied to Statistics
 p = 0.05 (average; reject the null hypothesis)
 p=0.01 (more statistically powerful)
 p = 0.001 (most statistically significant)

Student's ttest
 most popular statistical procedure for testing the difference between two groups of subjects
 applied to betweensubjects designs
 parametric test (assume normal distribution)
 used if # of subjects is less than 30

Ztest
 parametric test (assumes normal distribution)
 used when the # of subjects is greater than 30

ANOVA
 analysis of variance
 allows groups of different variables to be compared at the same time
 used at interval or ratio levels of measurement
 participants randomly samples
 assume normal distribution (parametric)
 populations have equal variance

ANOVA Classification
 one way (simple): one grouping factor; determine if the variance is because of sampling error or the treatment effect
 multifactor (complex): analyzing interaction of factors and separate effects of each factor

Format for Final Paper
 Table of Contents
 Abstract
 Chpt 1: Intro
 Chpt 2: Research Methods
 Chpt 3: Results/Data
 Chpt 4: Discussion
 Chpt 5: References
 Chpt 6: Appendix

Rating Scales vs Likert Scales
Rating Scale: assign value to an object
Likert Scale: 5 to 7 point rating scale that determines the extent to which someone agrees or disagrees with a statement

