Research Methods Exam 1
Levels of Evidence
(Primarily used in EBP) Ia Ib IIa IIb III IV
The scientific method
- "A method or procedure... consisting in systematic observation, measurement, and experiment, and the formulation, testing, and modification of hypothesis" - Oxford Dictionary - Process by which scientists, collectively and over time, endeavor to construct an accurate representation of the world. (Wolfs, 2013) o Accurate = reliable, consistent, non-arbitrary - we do not want results to be random/subjective - **The Scientific Method attempts to minimize the influence of bias or prejudice in the experimenter when testing a hypothesis or theory
Research
- "A systematic investigation (i.e. the gathering and analysis of information) designed to develop or contribute to generalizable knowledge" -45 CFR 46.102d - The object of research is to "Extend human knowledge of the physical, biological, or social world beyond what is already known" -National Academy of Sciences - As long as you are contributing something to the body of knowledge that is out there, it is still good and it is still considered research - Qualitative Study - observe
Science
- "The use of evidence to construct testable explanation and prediction of natural phenomena, as well as the knowledge generated through this process" - National Academy of Sciences - "... a branch of knowledge or study dealing with a body of facts or truths systematically arranged and showing the operating of general laws; systematic knowledge of the physical or material world gained through observation and experimentation" - dictionary.com - "...the state of knowing: knowledge as distinguished from ignorance or misunderstanding" -Merriam-Webster o Three very different definitions of science; but what can be used as a definition as a whole
Other citations to be aware of
- AMA (American Medical Association - MLA (Modern Language Association) Chicago/Turabian
What is APA
- American Psychological Association - Most commonly used format within the social sciences Currently 6 editions, wth the most recent edition being the Publication manual of the American Psychological Associateion - If date comes directly after an authors name
Titles - General rules for apa
- Book titles, periodicals (including journals), web sites, dissertations/theses, reports/technical papers, and works of art are always italicized - The title of a journal article is NOT italicized - Example: o Journal: Aphasiology o Article Title: Semantic typicality effects in acquired dyslexia: Evidence for semantic impairment in deep dyslexia
Protection of Human Participants
- Brief Timeline: o Nuremburg Code (1949) ♣ Post WWII - after public realized WWII experiments ♣ Set laws to protect US human participants o Declaration of Helsinki (1964) ♣ Published by the World Medical Association ♣ Based on Nuremburg Code but increased anti with medical research on humans o Belmont Report (1979) ♣ Helped create the IRB within the US ♣ Research used to be sealed - didn't need to tell the public human research needs to be public ♣ Created the IRB • IRB weighs risks and benefits to see if it will actually contribute to science • *IRB applies only to individuals that participate in research • If you are just doing treatment in a clinic, you do not have to write an IRB ♣ Summarized all ethical principles and guidelines in the US and protects what all researchers need to abide by o 1974 - National Research Act
Ethical Issues with Treatment Methods
- Concerns with informed consent o Do patients really understand what they are signing? o Do they understand randomization? o Signee's state of mind @time of consent - No-tx control groups o Tx denial acceptable when no other tx is available, OR if the control group is receiving an existing/standard tx o Ex: Cancer- people still need to be given original treatment and then another group will get new treatment - Issues with Placebo Control Groups - Randomization & recruitment o Problem with SLP is repeated use of same participants ♣ You cannot use clients for Ewing Clinic for every study...
Contextual Characteristics
- Conditions in which an intervention is embedded or the experimental arrangements that are key to implementing an experiment may restrict the findings to those conditions/arrangements only - Subjects who are aware of the study, or who guess the purpose of the study, may be influenced to answer/participate in a certain way - Would the same results be obtained if the subjects didn't know they were being studied or didn't guess the purpose of the study? o When someone can guess purpose, it is going to affect how they are participating
Ethical Justification of Treatment Evaluation
- Dilemma is whether to experiment or not experiment with different tx procedures - Is the treatment necessary? o 3 assumptions: ♣ Effective procedure already exists and effects are objectively documented ♣ No room to improve efficacy of documented procedure ♣ No need to find a more objective procedure - Is the treatment ethically justified? - Are treatment and treatment research different? In SLP, there is no treatment that says this is how you treat all people.
Qualitative Research
- Emphasizes data collection in a "natural" setting - Often observatory - Uses a variety of descriptive and interpretive methods that remain more flexible in application, and that allows for the discovery of new leads to knowledge as the data merge - Used widely in the social sciences o Sociology, educational research - Ex: observe children with high literacy skills get read to more frequently - More flexible in looking at overall question you are asking
Dissemination of Research Findings
- Ethical responsibility of researchers - Ethical principles are meant to reduce the undesirable side effects of scientific research and thereby increase the potential benefit to humankind - Our responsibility to hold our fellow researchers and clinicians to a higher standard
IIa
- Evidence obtained from a well-designed control study without randomization Ex: Subjects with muscle tension dysphonia - look at cross section area of hyoid (use a chunk of normal healthy speakers to be in the control group)
Effects of Science on Society
- Generally speaking, scientific research is for the good of society o However, every good effect has a bad effect as well o Particularly true within the Natural Sciences - Examples: atomic research, genetic research, chemical research, etc. o Atomic Research nuclear medicine (MRI), also atomic bomb o Genetic Research human genome (cloning) o Chemical Research drugs (insecticides)
Ia
- Highest level of credibility - Well-designed meta-analysis of more than one randomized controlled study (or a systematic review) o Meta-Analysis: statistically combines the results of multiple scientific studies ♣ E.g. In medical journals, 1 or 2 physicians find all possible randomized control studies on a drug and statistically analyze data from all of these studies o Systematic Review: A summary of evidence conducted by one or more experts that identifies, synthesizes, and appraises studies to answer a particular question or draw a conclusion about a specific area/topic ♣ 10-page review on state of how we are addressing literacy in schools ♣ Structured like a lit review
4 Subtypes of Construct Validity
- Inadequate explication and operationalization of constructs - Single operations and narrow stimulus sampling - Experimenter expectancies - Cues associated with the experimental Situation
Quantitative Research
- Involves a systematic and highly disciplined approach to problem solving - Formalized tests and instrumentations are applied to precisely and objectively specify data characteristics in numerical terms - Ultimate goal is to show that the hypothesis is true or false - True Experimental: random assignment, active manipulation, experimental group compared to Tx group o Biology study, animal studied - Quasi-Experimental: assigned to groups based on preexisting conditions, constructed control group or no control group, used when true experimental is not possible o Assigned to groups based on preexisting conditions o May not have a control group o Most of the research is SLP and Audiology
Articles
- Library database with a Digital Object Identifier (DOI) o Author (date). Article title. Journal Name, V(#), 12-23. doi: xxxxxxxxxxxxxxxxxxxxx - Library database without a DOI o Author (date). Article title. Journal Name, V(#), 12-34. Retrieved from (insert website). ♣ Website can be journal homepage if full-text available ♣ If not, use database web address
5 subtypes of Statistical Conclusion Validity
- Low statistical power - Violated statistical assumption - Fishing and error rate - Reliability of measures - Reliability of treatment implementation
IV
- Lowest level of evidence/credibility - Includes expert committee reports, consensus conference, clinical experience of respected authorities - At times, may be the only source available o Particularly with unique cases that have not been published yet and are being presented at a particular conference
IIb
- Majority of SLP studies - Evidence obtained from a well-designed quasi-experimental study o Assigned to groups based on preexisting conditions, constructed control group or no control group, used when true experimental is not possible
Protection of Animal Subjects
- Not common in SLP - Have comparable ethical guidelines to human research - Institutional Animal Care and Use Committee (IACUC) o How will they be treated, euthanized, etc. in humane manner?
Extraneous Variables
- Nuisance factors unrelated to the DV or IV that might exert unwanted influences on the outcomes of a research study o You always want to identify as many extraneous variables as you can before you start your study, so you can control for them (if possible). o "Problems are best solved in advance instead of second-guessing what might have been"
Consequences of Ethical Constraints
- Particularly when referring to treatment research - One always needs to consider the conditions under which data was collected when making interpretations - Informed consent and Voluntary Participation o When subjects are given the full details of a study, a biasing effect may be created o Some research questions can't be answered unless the study purpose is concealed ♣ Ethical = full disclosure ♣ Aspects may be concealed only when the study itself does not produce any negative effects for the subjects, or there is no other way to study the phenomenon o Research is voluntary, and the issue with autonomy is that all participants are self-selected ♣ Ultimately affects randomization
Fraud in Scientific Research
- Personal conduct of scientists o Should be people of integrity - Misconduct in Science: "...fabrication, falsification, or plagiarism in proposing, performing, or reporting research. Misconduct in science does not include errors of judgement; errors in the recording; selection of data; or misconduct unrelated to the research process" - Also includes data distortions, deceptive data analysis, deceptive data reporting, selective suppression or deletion of data, etc. - Can stem from pressure to produce certain types of data... "publish or perish"
Statistical Conclusion Validity
- Pertains to the approximate accuracy of inferences about the covariation of variables based on the specific research operations and statistical tests used in an investigation - Major concern for researchers is variability o Degree to which data are indicative of intended systematic influences as opposed to uncontrolled chance factors Type I & Type II errors
General Research Types:
- Qualitative Research - Quantitative Research o True Experimental o Quasi-experimental
Literature Reviews Goal
- Read broadly enough to uncover present trends and controversies pertaining to the research topic but also selectively with the aim of discovering gaps in existing knowledge and identifying problems needing further study
Originating a question
- Start broad - Have a good textbook understanding of your topic o Ex: Child Language (what are the stages of development, what constitutes a disorder, what types of therapies are seen as an SLP) o Ex: Voice (physics behind vocal fold dynamics, physiology, different disorders) - Should be interesting (to you) and should make sense (to you) o If a compelling rationale for conducting a research study cannot be established, the results are likely (to be viewed as) unimportant - Identify current trends o Current = roughly 7-10 years o Depends on topic area
Volume and Issue Numbers
- The volume number is italicized; the issue number is not. o Example: 15(3)
When is the Scientific method not applicable?
- Unable to isolate the phenomenon of interest o Ex: larynx - cannot take it out of a person to look at it and put it back in - Unable to eliminate or account for extraneous factors o One of the hardest parts of designing a research study, is the attempt to control for all variables that will impact the study - Unable to test after making controlled changes o If you change a part of your study, you cannot control for things afterward * These can be based on ethical reasons, or situational reasons
Rationale for a Question
- Underlying reason for asking the question o Explain the thinking processes behind the research ♣ Will the answer to the question...? • ...help confirm/refute/extend previous findings? • ...be generalizable to other settings/circumstances? • ...provide new info of current clinical significance or scientific importance? • ...help advance the development of theory and new research?
Page Numbers
- Use "p." and "pp." (single and multiple) to indicate page numbers when citing newspaper articles and book chapters. - Omit them when citing journal and magazine articles
Ib
- Well designed randomized controlled study o Ex: control group = no tx, Tx=tx ♣ Everyone in treatment group gets treatment, everyone in control group gets no treatment - everyone is completely randomized
III
- Well-designed non-experimental studies o Correlational and/or case studies o Retrospective or Ex Post Facto research ♣ Search of the cause of an event without the IV ♣ Ideal for conducting research when it is not possible or acceptable to manipulate the characteristics of human participants ♣ Like "Research in reverse" ♣ Retrospective- information already out there and you are using them; l • Ex: hospital records related to hospital induced aspiration pneumonia back in 2012 • Research in reverse - do not manipulate anything! • Ex: children read to on a frequent basis (look at standardized testing scores for that particular school)
Feasibility of Answering a Question
- What means are available to accomplish the intended task - Is the question/hypothesis stated in a testable form? o IV/DV - Methodological constraints? - Estimated cost? - Difficulty of project or time limitations? - Ethical Issues?
Internal Validity
- When the selected IV produces acceptable changes in the DV in the absence of other potential variables, or the extent to which an investigation rules out alternative explanations for observed results - Certainty that researchers infer that a relationship between two variables is causal 6 subtypes
Quasi-Experimental
- assigned to groups based on preexisting conditions, constructed control group or no control group, used when true experimental is not possible o Assigned to groups based on preexisting conditions o May not have a control group o Most of the research is SLP and Audiology
True Experimental
- random assignment, active manipulation, experimental group compared to Tx group
8 Fundamental Ethical Issues that Affect Research
1. Fraud in Scientific Research 2. Effects of Science on Society 3. Ethics of Treatment Evaluation 4. Protection of Human Participants 5. Consequences of Ethical Constraints 6. Ethical Issues with Treatment Methods 7. (Ethical Issues Relative to Animal Subjects) 8. Dissemination of Research Findings
4 Main Elements of the Scientific Method
1. Observation and description of a phenomenon or group of phenomena 2. Formulation of a hypothesis to explain the phenomena 3. Use of hypothesis to predict the existence of other phenomena, or to predict quantitatively the results of new observations 4. Performance of experimental tests of the predictions by several independent experimenters and properly performed experiments (Wolfs, 2013)
Formulation of a Problem:
1. Originating a question 2. Developing a rationale or statement of importance 3. Determining the feasibility for answering it
4 subtypes of External Validity
1. Sample Characteristics 2. Stimulus Characteristics 3. Contextual Characteristics 4. Assessment Characteristics
3 Levels of Understanding Lit Reviews & Levels of Evidence
1. What? o Intended goal often to develop taxonomies o Important in setting up mutually exclusive dx criteria for identifying disorders according to clusters of certain signs or symptoms - Ex: Swallowing (dysphagia has not been discovered); suddenly, many patients have problems swallowing thin liquids (coughing on thin liquids) o Trying to classify this group of patients 2. Where? o Might pertain to the direction of associated changes in one variable in relation to another? - Ex: Swallowing; thin liquids cause coughing, thicker liquids do not cause coughing 3. Why? o Characterizing the various features of variables and their relationship to others; discerning causal relationships between variables (Research and science) - Ex: Swallowing; thin liquids cause people to cough; thicker cause people not to cough. Why are they coughing on the thin liquids? Do they have a esophageal sphincter malfunction?
Type II errors
Accepting null hypothesis when it is false - False Negative - not pregnant when it's obvious you are pregnant - Guilty person is free
Regulating Body of Levels of Evidence
Agency of Healthcare Research & Quality (AHRQ)
What is Validity
Aka validity of experimental operations - The confidence with which the experimenter's claim of a cause-effect relation can be accepted... experimental results are valid when the observed cause of an event is not mistaken - The ultimate goal of scientific inquiry in any discipline is causal explanation o Involves: ♣ Identify what variables are related ♣ The nature of the specific mechanisms/processes involved ♣ The extent to which relationships can be generalized across populations, settings, and conditions
5 Major Aspects of Phenomena in SLP
Articulation Language Fluency Voice Swallowing
Experimenter Expectancies
Construct Validity - If the primary investigator has certain expectations of a method or treatment, I may confound how they interpret findings - May be observed as a change in the DV - Ex: you create a new treatment method for /r/. Enthusiasm with new therapy and no enthusiasm with old therapy will impact the DV. Will impact how the kid responds to the particular therapy
Cues Associated with the Experimental Situation
Construct Validity - May refer to inadvertent cues given by the researcher, or incidental information given to the subjects from outside sources o ... that contribute to the study results - Threat to construct validity when the cues represent plausible, alternative explanations for the causal mechanisms that are presumed to account for the experimental findings o The cues can lead to changes in the DV - Always be mindful of giving fewer cues to participants - can skew particular subject to do something a certain way
Inadequate explication of operationalization of constructs
Construct Validity - Pertains to the "fit" Between conceptual and operational definitions used by the researcher - Threat to construct validity when the commonly accepted definition (of X) doesn't match or fit the researcher's definition (of X) - We should all say tomato
Single Operations and Narrow Stimulus Sampling
Construct Validity - Refers to using a single measure at a single point in time as the sole indicator of something (aka a particular construct) - Using multiple indicators that represent different methods of measurement allows for separation of the variation due to the construct from the variation associated with the measurement method
Stimulus Characteristics
External Validity - Characteristics of the setting, experimenters, or how the stimuli are presented in an experiment - Threat to external validity when the specific features of a given intervention restrict generalization of the findings (to those conditions only) - Examples: using only 1 experimenter to implement intervention, using one specific stimuli in the same order, using a specific setting which may be different than other settings o If you are only doing your testing in clinic vs testing in classroom o Medical vs. home setting can sway results
Sample Characteristics
External Validity - Extent to which findings may apply to persons who vary in age, race, ethnic background, education, or other characteristics who comprised the research sample - Degree that a sample of subjects truly reflects the target population to whom we wish to extrapolate our findings - "to" vs "Across" a population o Mostly looking to compare our sample to a population o Across a population reflects that population and all subgroups in it ♣ Ex: ASD - only comparable to low functioning ASD, you cannot apply to all ASD - Becomes a threat to external validity when population is different than what you are comparing population to
Assessment Characteristics
External Validity - How the DV is assessed may influence subject response o Restricts external validity to similar assessment modes - The assessment instrument may cause the subjects to be aware that their performance is being evaluated, which may affect their responses - Subjects may also become sensitized to the assessment instrument (e.g. pre-post testing, or multiple baselines)
6 Subtypes of Internal Validity
History Maturation Instrumentation Selection Attrition Ambiguity about Direction of Causal Inference
Research Variables
Independent Variables & Dependent Variables
Four Types of Research Validity
Internal Validity External Validity Construct Validity Statistical Conclusion Validity
Selection
Internal Validity ♣ Also referred to as "subject selection bias" ♣ Threat to internal validity when the effect of an intervention may be the result of the different types of people who choose/choose not to participate in research ♣ Addressed via random sampling and/or random assignment • Random assignment - match people in experimental group to people in control group ♣ Doesn't really apply to single-subject design (not making group comparisons)
Maturation
Internal Validity ♣ Any biological or unidentifiable changes that take place in participants as a result of time (maturity - growth along a lifespan) ♣ Threat to IV when an observed effect may be due to participants' growing older, wiser, stronger, or better adjusted between a pretest and posttest when this maturational process is not the target of the investigation • Ex: Interested in MLU for children that are 2. Hypothesis Is when a child is 3, the MLU will increase. Growth obviously occurs in the year. • Ex: Anatomy and physiology research- VF length measurements (14 to a 30-year-old)
History
Internal Validity ♣ How a participants' life events are totally or partially responsible for the changes recorded in the DV after the introduction of the IV ♣ Events which are common to all subjects in their everyday lives or unplanned circumstances occurring during the process of implementing an experimental procedure that may plausibly represent rival, competing explanations of the results ♣ Threat to internal validity when the event impinges on the experimental situation in some way and influences scores/measures on the DV • Ex: survey on impressions of disability services; someone with experience with people with special needs is going to be different from someone with no experience (Life experience can effect DV is hypothesis is clinicians have a higher empathy for individuals with a disability)
Attrition
Internal Validity ♣ When participants are lost during the course of study or experiment (withdraw, drop out, data goes missing) ♣ Threat to internal validity when the effect of an intervention may be due to systematic differences associated with the characteristics of individuals who withdrew during the course of the experiment ♣ Doesn't typically apply to single-subject designs • You are not using statistics to determine sample size ♣ Always use 20% more people for sample size than you think you need
Independent Varialbes
Manipulated, assigned, or grouped by research in order to examine their effect on the depdnet variable ♣ Manipulated vs. Non-manipulated • Non-manipulated IV's are not controlled by the researcher but preexist as states or conditions within the participants of a study o Ex: in prof's dissertation, she did not control for gender (non-manipulated IV) • Only through active manipulation of the IV can an outcome be casually related to the influence of that manipulation
Measurement Variables
Qualitative Variables & Quantitative Variables
External Validity
Refers to issue of generalization o Extent to which the investigator can extend or generalize the results to other situation, participants, experimenters, service context, etc. o Random selection of research subjects affords the strongest warrant that results will generalize to other individuals in that population ♣ Can be problematic with single-subject designs o Perfect example of generalization: when empirical findings of several studies are consistent across various types of subjects, settings, and other conditions ♣ Ex: subjects can be college students, school aged, ASD ♣ Settings: clinic, school, home, hospital ♣ Other conditions: researchers doing exact same thing getting similar results, across different cultures?
Construct Validity
Refers to the specific causal factors or mechanisms that are responsible for the observed change in the DV o Also used to describe the extent to which the results obtained from a study converge with those found in other investigations of the same phenomenon to support a particular theory o Threats to construct validity are confounds that call into question the researchers interpretation regarding the specific factors that account for the study findings o Often very difficult to establish... o Operational Definition: how are we defining a specific measurement ♣ Should be the same across the board so all researchers know
Reliability of Measures
Statistical Conclusion Validity - Are the instruments used to measure the variables reliable? o A large error in measuring the DV may increase the likelihood that mean differences between groups doesn't reflect true differences between populations o Or... observed relationship between two variables is not a true relationship o Inter-rater reliability and intra-rater reliability ♣ When there are raters/people involved ♣ Inter-rater: between two people (you measure, then your advisor measures to see if measures are equal) • Different measures mean one person is doing something wrong • If it is small enough, other rater only needs to do a certain amount of images ♣ Intra-rater: rater within themselves - how are you consistently rating something over and over again • Repeat 20% of whatever your sample is • Ex: 100 images - 20 are repeated to see if you measured the same images in the same manner
Reliability of Treatment Implementation
Statistical Conclusion Validity - Is there a lack of standardization for implementing a treatment or intervention? - If multiple researchers are giving a treatment/intervention, are they all doing it consistently the same way? o Studies usually implement pre-study training o Everyone should be doing everything the same way - implement treatment, collect data, etc. in same exact way each time
Violated Statistical Power
Statistical Conclusion Validity - Most statistical tests require certain assumptions to be met in order to be conclusive o Example: any statistical test involving group means comparisons assume that the groups have equal variance ♣ Parametric vs. non-parametric tests • If there is variance between groups, or they are not normally distributed, you have to use non-parametric Assumptions must be met in order for validity to be met
Fishing and Error Rate Problem
Statistical Conclusion Validity - Performing multiple statistical tests without correcting for the number of comparisons increases the probability of concluding that covariation exists o Results in a false-positive, or Type I error o Bonferroni correction is most common way to address multiple comparisons - Any time you do a statistical test and you do it multiple times, you have to do a correction o Level of significance gets cut
Low Statistical Power
Statistical Conclusion Validity - The "power" of a statistical test is the probability that it will detect a significant relationship between two variables, and ultimately reject the null hypothesis - Power consists of sample size, effect size, and alpha level - Low power may result in an invalid statistical decision o .05 - standard alpha level ♣ Statistically significant ♣ Greater than .05, results are not significant - When you have low power, that can result in an invalid statistical significance
Additional Subtype of Internal Validity
Statistical Regression o Return from an extreme point to an average level ♣ Can refer to medical condition or test scores, for example. o It is a threat to internal validity if the clinical condition is at its worst at the beginning of the experiment, but returns to average level as the study progresses ♣ If you are looking at general medical diagnosis, if you do not take into consideration the person's health at the beginning of the experiment, the person will obviously get better and determine results (FLU) o Can give false-impression of positive effects * ♣ Do not want positive effects to be due to individual issues
Parts of a research article
Title Author Abstract Key Words Conflict of Interest/Financial Support Introduction Method (subjects, equipment, interventions/study procedures, data analysis) Results Discussion Conclusion Acknowledgements
Why do we need citations
To avoid plagarism
Systematic Review
Summary of evidence conducted by one or more experts that identifies, synthesizes, and appraises studies to answer a particular question or draw a conclusion about a specific area/topic
How is research different from other forms of discovery
The Scientific Method
Plagarism
an act or instance of using or closely imitating the language and thoughts of another author without authorization and the representation of that author's work as one's own, as by not crediting the original author; literary theft
Impact Factor
average number of citations received per paper published during the two preceding years - SLP is low
Qualitative Variables
can be grouped or placed into categories; pertain to how people or things are placed together according to one or more attributes ♣ "Either/or" categories of observation ♣ Dichotomous (2 levels) - male/female ♣ Polyotomous (more than 2 levels)
3 Broad Classification of variables
o According to their measurement properties o How they are used in various types of research o The degree to which they exert an extraneous or confounding influence on the outcome of an experiment
Dependent Variables
o Dependent Variables: observed effect or outcome assumed to be associated with/caused by the independent variable ♣ Ex: people phonate on /e/ and /æ/ (Independent Variables) - dependent variables were tongue height, tongue advancement ♣ Ex: independent variable: drug; dependent variable: about of vocal tremor when taking drug
Search Engines
o Medical: ♣ PubMed (National Library of Medicine) ♣ Medline/EBSCO ♣ Scopus ♣ Science Direct ♣ CINAHL (interdisciplinary research - speech and PT/OT) o Educational (phonology, literacy, child language) ♣ Academic Onefile ♣ Academic Search Premier ♣ ProQuest ♣ Psychology/PsychInfo & ERIC
History of a research problem
o Theoretical and practical importance of research problem ♣ Deficiencies in current knowledge • Purpose of present research
A solid literature review should allow you to:
o Verify that you have chosen an appropriate topic o Define and refine aspects of your own study o Begin to estimate the resources your study will acquire o Establish the importance of your topic and justify the effort you will expend
What needs to be credited?
o Words or ideas presented in a magazine, book, newspaper, song, TV program, movie, Web page, computer program, letter, advertisement, or any other medium o Information you gain through interviewing or conversing with another person, face to face, over the phone, or in writing ♣ Excerpt from interview... o When you copy the exact words or a unique phrase o When you reprint any diagrams, illustrations, charts, pictures, or other visual materials o When you reuse or repost any electronically available media, including images, audio, video, or other media o Bottom Line: if you didn't think it/say it/write it, cite it!
Quantitative Variables
o are numerical; represent a measurable quantity; ordered/ranked according to magnitude, express extent to which objects differ in degree, not in kind ♣ Height, age, weight ♣ MLU, fundamental frequency (voice research)
Type I errors
rejecting the null hypothesis when it is true -False Positive - pregnant when you're not -Innocent person goes to jail
Meta-Analysis
statistically combines the results of multiple scientific studies
Ethics
system of moral principles; the rules of conduct recognized in respect to a particular class of human actions or a particular group, etc.
Phenomenon
tend to be more bored Variables are aspects of the event
Variables
tend to be more specific - narrowly defined aspects of an event that can be measured and manipulated - When we talk about variables, we are talking about the noun; the thing that changes A variable is... "any attribute or property in which organism (objects, events, people) are observed to vary" o Measurement Variables o Research Variables o Extraneous Variables
3 Kinds of IV's in SLP
• 1. Normal Behaviors (Event) • 2. Disordered Behaviors (event) o age appropriate • 3. Treatment Technique o manipulated, assigned, grouped o Ex: amount of biofeedback that is given (Ultra Sound study)
IRB
• IRB weighs risks and benefits to see if it will actually contribute to science • *IRB applies only to individuals that participate in research • If you are just doing treatment in a clinic, you do not have to write an IRB
Instrumentation
♣ A threat to internal validity when an effect, as measured on the DV, is due to systematic changes in the measuring instrument from pretest to posttest ♣ Instruments aren't just mechanical • Standardized Tests • Any pre-post testing material is considered an instrument
Deliberate vs accidental plagiarism
♣ Deliberate: do it and hope you don't get caught ♣ Accidental: some professors treat all the same; this is something else - may not realize they did it because they don't know the APA guidelines, by mistake and forgetting ♣ You must give credit where credit is due
Ambiguity about direction of causal inference
♣ Threat to internal validity when it is not possible to determine with certainty whether variable A causes variable B, or vice versa ♣ Correlation doesn't always imply causation ♣ Not usually a problem when the researcher can specify which variable comes first (Causal agent) and which variable follows (the effect) • Stress and anxiety - hard to separate these two ♣ More difficult in correlational studies - hard to say what is causing agent vs. effect
