6805 Textbook

Step One of The Scientific Method

Identifying a problem and defining it through background research.

Step two of The Scientific Method

Developing a specific hypothesis or question to investigate.

Step three of The Scientific Method

Planning a set of procedures for testing the hypothesis or answering the question.

Step four of The Scientific Method

Collecting data using those procedures.

Step five of The Scientific Method

Analyzing the data.

Step six of The Scientific Method

Making a decision regarding the viability of the hypothesis or the answer to the question.

Scientific Curiosity

•  Satisfying the need to understand human communication physiology and the causes of disorders.

Guiding Clinical Practice

 Moving away from sole reliance on expert opinion toward evidence-based practice (EBP).

Program Evaluation:

Localized research used to evaluate the quality of services against national standards or consumer satisfaction.

Public Policy:

• Influencing the allocation of resources by demonstrating the long-term impact of programs (e.g., the benefits of early childhood education).

Cost-Effectiveness and Cost-Benefit Research

•  Cost-effectiveness research compares the cost of a program to the produced outcomes. Cost-benefit research looks at the cost of a service relative to its impact on long-term costs later in life.

Empirical Research

Involves the collection of new information or data through the observation and measurement of behavior or physical properties.

Nonempirical Research:

Makes use of existing information instead of gathering new data. This includes literature reviews, theoretical analyses, and systematic reviews.

Quantitative Research:

 Relates to numerical information, such as frequency counts, measures of size, or physical properties.

Qualitative Research:

Involves verbal information, such as highly detailed descriptions of behavior or direct quotes from participants

Experimental Research

 Researchers manipulate or control one or more factors (independent variables) to see the effect on an outcome.

True Experiment:

Includes both the manipulation of variables and the random assignment of participants to groups.

Quasi-Experimental

•  Includes manipulation but lacks random assignment (e.g., using pre-existing classrooms).

Nonexperimental Research:

•  Researchers investigate existing conditions without manipulation. This includes descriptive studies, case studies, surveys, and correlation studies.

Case-Control Studies:

 A form of nonexperimental research where groups are compared based on pre-existing differences (e.g., comparing persons with hearing loss to those with normal hearing).

Group Research

  Participants in one group experience the experimental condition while another group serves as a control. Results are aggregated and reported for the group as a whole.

Single-Subject Research:

This is experimental in nature. A participant experiences both the experimental and control conditions, and results are reported separately for each individual participant.

Independent Variable:

The characteristic or manipulation the researcher wants to study (the "input"). In experiments, it is the manipulated experience; in nonexperimental research, it is an existing characteristic like age or the presence of a disorder.

Dependent Variable

The measures used to determine the outcomes of the research (the "output").

Respect for Persons

 Recognizing the autonomy of individuals. Potential participants must be able to make informed, voluntary decisions without coercion or excessive enticement.

Beneficence

 The obligation to protect participant well-being. Researchers must maximize possible benefits and minimize possible harms ("do no harm").

Justice

 The requirement for equity in research. The burdens and benefits of research should be shared fairly. Researchers must have a sound scientific basis for including or excluding specific groups.The "Common Rule" (Federal Policy for the Protection of Human Subjects) provides specific regulations for agencies receiving federal funds, ensuring voluntary participation, confidentiality, and the right to discontinue.

Federal law identifies specific groups that are vulnerable to exploitation or may have an impaired capacity for consent:

• Prisoners.

• Children.

• Pregnant women and fetuses.

• Persons with impaired consent capacity

Tuskegee Syphilis Study (1932–1972):

• This study investigated untreated syphilis in African American men. It violated all three Belmont principles: participants were misled ("bad blood"), penicillin was withheld as an effective treatment for decades (violating beneficence), and a vulnerable, impoverished population was exploited (violating justice).

Willowbrook State School Hepatitis Study (1955–1970s):

•  Researchers deliberately infected children with disabilities with hepatitis. Ethical concerns included incomplete or misleading consent forms and coercive offers (earlier institutional placement contingent on research participation). Critics argued that a known treatment (gamma globulin) was withheld.

Tudor Study ("Monster Study," 1939):

Conducted at the University of Iowa, this study used negative verbal labeling on orphaned children to see if it induced stuttering. It violated the principle of justice by using a "sample of convenience" (orphans) and beneficence by exposing children to psychological risk without potential benefit.

Institutional Review Boards (IRB)

Universities and medical centers are required to have an IRB to evaluate proposed studies before data collection. The IRB conducts a risk/benefit analysis to ensure the study adheres to ethical guidelines and protects participants.

.Scientific Misconduct and Integrity

 Research integrity involves the responsible conduct of research, including:

Accuracy in Reporting

Avoiding deliberate misrepresentation, fabrication, or falsification of data.

Attribution of Ideas:

• Referencing sources correctly to avoid plagiarism.

Treatment Fidelity:

• Strategies used to ensure a study is executed accurately and consistently according to the protocol.

Credit for Intellectual Effort

Assigning authorship based on substantive contributions to the conceptualization, design, or analysis of the research.

Clinician-Researcher:

Balancing the role of treating a client with the role of investigating a new, unproven treatment.

Product Developer:

• Conducting research on a product one intends to market.

Instructor-Researcher

•  Recruiting one’s own students for research, which can be perceived as coercive.

.HIPAA and PHI

The Health Insurance Portability and Accountability Act (HIPAA) "Privacy Rule" protects "Protected Health Information" (PHI). Clinician-investigators must comply with HIPAA if they work for or access records from "covered organizations" (health plans, clearinghouses, or electronic health care providers).

De-identified Information:

Removing 18 specific categories of identifiers (name, address, social security numbers, facial images, etc.).

Limited Data Sets:

Removing 16 categories but retaining some geographic data (requires a data use agreement).

Written Authorization

•   Obtaining specific permission from individuals.

Waiver of Authorization

  Obtaining a waiver from an IRB or Privacy Board

Abstract

• A brief overview of a research article's contents; "structured abstracts" include specific headings like Purpose, Method, Results, and Conclusion.

Beneficence

• An ethical principle requiring researchers to protect participant well-being and maximize benefits while minimizing harms.

Case-Control Study:

•  A nonexperimental study comparing groups with pre-existing differences, such as age or clinical diagnosis.

Clinician-Investigator

 A professional whose primary responsibility is clinical service but who also engages in systematic research

Common Rule

•  The "Federal Policy for the Protection of Human Subjects," which outlines essential protections for human participants in federally funded research.

Conflict of Interest

 A situation where the obligations of one professional role (e.g., clinician or instructor) conflict with the responsibilities of another role (researcher).

Cost-Benefit Research

Research evaluating a program's cost relative to its long-term financial impact on society.

Cost-Effectiveness Research:

 Research evaluating the cost of a program relative to the specific outcomes produced.

Dependent Variable:

• The outcome measures or "outputs" used to determine the results of a study.

Empirical Research:

• Investigations involving the collection of new data through observation and measurement.

Evidence-Based Practice (EBP):

•   An approach to clinical decision-making that integrates the best available research evidence, professional expertise, and client/family considerations.

Independent Variable:

• The characteristic or experimental manipulation being studied (the "input").

Time constraints:

Some questions require that you gather information over a long period of time. One of the most impressive studies we have read was a study of the impact of participating in a preschool program on children's subsequent achievement in school and in life (Barnett, 2000). The researchers followed their participants from preschool into adulthood, spanning more than 15 years. While highly valuable, such longitudinal designs are not practical to undertake in the typical time frame of a graduate program.

Required skills

Ensuring that the investigator possesses or can reasonably acquire the specific clinical, testing, or statistical skills necessary to execute the study.

Participant access:

Confirming access to an appropriate and sufficient group of participants representing the target population.

Equipment and facilities:

 Some research questions may be practical only in research facilities associated with major medical centers, such as those focusing on speech perception after cochlear implantation, neuroimaging studies, or genetic bases of disorders. These require access to specialized, highly expensive equipment and supportive personnel

Supportive personnel and mentorship

• Having access to supportive administrative staff and, crucially for developing investigators, a knowledgeable mentor in the interest area to guide the research.

. Extend to a new population:

Apply the study to a new age group, a different diagnostic group, or persons from a different linguistic or cultural background.

Apply different outcome measures:

Use measures from spontaneous speech if the original study used formal tests and/or measures from elicited speech samples, or perhaps use a newly developed measure.

Address questions of social validity:

Determine the functional impact of observed changes for the participants or their families (Foster & Mash, 1999; Strain et al., 2012).

Change the stimuli or presentation:

Change the stimuli used to elicit responses or the procedures for presenting stimuli to test whether findings are specific to certain conditions.

Alter the setting:

 Change the setting in which an evaluation or treatment occurs to provide either a more naturalistic setting or one that more closely matches real clinical situations.

Utilize updated instrumentation:

 Use up-to-date instrumentation in testing and analyzing participants' responses for approaches that might be outmoded

Descriptive Study

What is the consonant-vowel (CV) ratio for preschool children at ages 18 months, 24 months, 30 months, and 36 months?

Relational Study

What is the relationship between CV ratio and mean length of utterance (MLU) in preschool children from ages 18 months to 36 months? What is the relationship between scores on a new measure to assess recognition of speech in noise and performance on the Speech Perception in Noise (SPIN) test?

Difference Study—Nonexperimental

What is the difference in CV ratio between children with normal hearing and those with moderate-to-severe hearing loss at 24 months of age?

Difference Study—Experimental

What is the difference in CV ratio for 24-month-old children with expressive language delay who participated in an eight-week intervention program compared to 24-month-old children with expressive language delay who received no intervention? How do adults with hearing loss perform on a test of speech recognition when fitted with hearing aids that employ two different noise reduction algorithms?

Null Hypothesis (H₀

• Stated in the negative, it assumes that the study's results will yield no significant differences between groups and/or no significant relationships among variables. Statistical analysis is used to either reject the null hypothesis (finding it not viable) or fail to reject the null hypothesis (finding it viable). The null hypothesis serves as a strategy to maintain objectivity, assuming no effect until data indicate otherwise.

Research/Alternative Hypothesis (H₁):

Usually stated in a positive form, it represents the exact outcome the researchers expect to find based on theoretical models or previous literature.

P (Patient/Population):

A detailed description of the target group, minimally including the disorder (e.g., aphasia, sudden onset hearing loss), age range, severity, and time since onset.

I (Intervention):

•  The specific clinical procedure, technology, or treatment program of interest (e.g., naturalistic language intervention, bilateral cochlear implants).

C (Comparison)

•  The alternative intervention being evaluated, such as a traditional treatment approach or no treatment at all.

O (Outcome):

•  The desired clinical change, which could be specific (e.g., expressive vocabulary gains, reduced dysfluencies) or broad (e.g., quality of life, activities of daily living).

Face Validity

Based on an informal, subjective judgment of how well a test appears to accomplish its intended purpose. It relies on internal standards rather than empirical evidence. Although informal, establishing face validity with untrained consumers (clients) is vital for maintaining motivation and attention.

Content Validity:

 A formal, judgment-based approach where expert panels evaluate test items against a well-known model or theory. The experts assess how broadly the items sample the content domain, whether response modes are appropriate, and if various skill levels are represented. Tests are often refined based on this expert panel feedback

Criterion Validity:

An empirical approach that compares a new test directly to an existing standard test (the criterion). Two primary subtypes exist:

Concurrent Validity:

Both the new test and the criterion test are administered to participants at approximately the same time to establish equivalence.

Predictive Validity:

Test scores are used to predict performance in a future activity (e.g., academic or career success) through longitudinal studies where participants are reevaluated over time (often a year or more later).

Construct Validity

Determines whether patterns of performance on test items reflect the underlying theoretical model or constructs (meaningful associations of behavior) the test is built to measure. It is established by analyzing two specific relationships

 Convergence:

 Showing that a person performs in a highly similar way on test items or different tests designed to measure the same underlying content

Divergence:

 Showing that a person's scores differ significantly across items designed to measure distinct, separate behavioral constructs (e.g., separating motor skills from cognitive abilities).

Interobserver Reliability

The consistency of results achieved when two different examiners or observers use the same procedure to test the same individuals at about the same time. Correlation coefficients are computed between the two examiners' scores.

Test-Retest Reliability:

The consistency of results from one administration to another when the same examiner tests the same individuals at two different times. This is highly critical in treatment outcome studies to demonstrate that observed changes are due to treatment, not measurement instability.

Parallel-Forms Reliability:

Consistency of results when researchers construct and administer two equivalent versions of a test containing comparable items measuring the same behavioral construct to the same participants.

Split-Half Reliability (Internal Consistency)

Established by administering a test once, splitting the items into equivalent halves (such as odd- and even-numbered items), and correlating the scores. This measures how consistently the items tap into the same behavioral construct. While useful because it requires only a single administration, reliability may be underrepresented because test length is halved.

Evidence-Based Practice (EBP):

 Information literacy is critical for clinicians to find existing evidence to answer questions regarding the etiology, assessment, prognosis, or treatment of communication disorders

Original Research:

•  Researchers use these skills to identify gaps in current knowledge, define research problems, and ensure their studies yield valid and reliable results.

Definition of a Literature Search:

• This is a systematic process of finding information within a body of written work that pertains to a specific topic. This may include searching within CSD or related fields like education, medicine, neuroscience, or psychology.

Finding Clinical Answers:

Clinicians conduct searches to find specific evidence to support treatment decisions for individual clients.

Systematic Reviews:

• Authors of nonempirical works, such as theoretical discussions or systematic reviews, must conduct exhaustive searches to identify all existing evidence on a topic.

Identifying Research Problems:

• Reading existing literature helps researchers find specific questions or identify conflicting findings between studies that warrant further investigation.

Replication

•  Finding a similar existing study does not necessarily negate a new study. Replication is essential for a strong body of research, especially if older studies are outdated due to changes in population characteristics or new outcome measurement tools.

Identifying Methods and Measures:

Searches help researchers identify successful procedures, outcome measures (such as reaction time or activity and participation restrictions), and theories or models to justify their research ideas

Generating Background Information:

• Literature provides the necessary context, such as prevalence statistics and concept definitions, required for the introductory sections of research reports.

Truncation (*):

A symbol used to handle grammatical variations by searching the root of a word (e.g., stutter* retrieves stutter, stutterer, and stuttering).