Research Exam 2
Internal validity
all about did the experimental manipulation control for the variable speed control. Internal validity has to do with internally a lot of manipulation and control variables so that you could say this variable cause this effect with certainty and confidence.
What are the three(3) types of exploratory research?
Cohort Case Control Correlational
Dilemma of randomized control trials
Ethical dilemma. Who will receive treatment and who will not.
Describe a negative correlation?
When one variable increases, the other variable decreases
Directional Hypothesis
You're looking at whether one variable increases or decreases something else (a response, a behavior, a set of skills.)
Define research design
"Provides the glue that holds the research project together. A design is used to structure the research, to show how all of the major parts of the research project work together to try to address the central research questions" provides a structure to the to the research so the structure that provides a framework it provides a recipe of steps that you need to follow in order to be able to successfully complete whatever study you are trying to design
Benefits of quantitative research
***1. allows for better interpretation what what methodology should I use.*** 2. Can create graphs bar graphs visuals (get my research questions can be answered numerically and can I display my data in a visual way that leads to better interpretation) 3. Quantitative Data 4. Policy makers/ administrators use quantitative data as decision makers (funding for Early Intervention, etc)
Steps in Hypothesis Testing
***Usually in the form of a Null Hypothesis Statement: State the hypothesis*** Select appropriate test statistics that you'll run (Most powerful are parametric tests like t-test or Pearson Correlation) Establish a level of significance: (Important b/c once you establish your level of significance, if the number falls below it's not significant, if the number falls above, it's significant) Helps determine if you will accept or reject the Null Hypothesis. Little p is the level of significance, ***usually 0.05 or less. Researcher is the one who has to establish the p value (level of significance).Is there a difference or is there no difference?*** Choose a sample size. Limitations: Having too small of a sample size because it's not representative of the population at whole, decreasing significance. ***The size of your sample matters! Smaller sample sizes are harder to generalize the results.*** Select a 1 or 2 tailed test Compute test statistics using t-or z-statistics on the ***SPSS, it also calculates degrees of freedom.***
Importance of research questions
-They provide the guide of your research. -Questions can be found in articles you read, or from class, clinical experience, something of interest to you, something of curiosity to you. Questions try to identify gaps in knowledge in a topic. -Questions can lead to the production of new knowledge (original science). -Questions can verify existing knowledge (replicative data) -Questions can lead to new questions.
List the four (4) types/categories of research.
4 categories of research: Experimental: Randomized Controlled, Quasi-Experimental, Sequential, Single Subject Design & Meta-Analysis Studies Descriptive: Cross sectional, longitudinal, Semi-Longitudinal, Historical, Case, Secondary Analysis & Evaluation Studies Exploratory: Cohort, Case Control & Correlational Studies Survey (Can be a category of research but can also be an instrument and can cut across experimental, exploratory and descriptive.)
Define meta-analysis
A statistical technique for combining findings from independent studies Uses previously gathered data Data are re-analyzed from the compiled data Often used to assess the clinical effectiveness of interventions/treatments Provides a precise estimate of treatment effect
Define experimental
A study in which a treatment, procedure, or program is introduced and a result or outcome is observed
Provide a definition for longitudinal study?
A study that analyzes one group over a long period of time.
ABAB design what does the first A stand for? And the first B?
A= non-treatment phase (baseline data) B= treatment phase
Scatterplots interpreting
All Data points are near the line. Whether positive or negative correlation depends on the slope of the line (Decrease in X, causes an Increase in Y, or a Decrease in Y causes and Increase in x)
Define hypothesis
An educated guess, you want to prove or disprove the hypothesis in order to contribute to the knowledge base.
Explain a longitudinal study.
Analyze one group over a long period of time Example - group of 1 year old children with fluency disorder until they turn 5 year olds
Types of questions answered in a descriptive design?
Answers "what is" and "what was" type questions
Example of a case study (2)
Cases are chosen b/c people in the study have a particular trait/disorder. Control participants are chosen b/c they do NOT have the trait/disorder. Used to examine the effects/impacts of that trait/disorder on the individual. Example: 1. Children with disfluency disorders-Case & Children with no disfluency disorder-Control (example she SPOKE ABOUT) or 2. People w/ Alzheimer's dementia-Case & People with no Alzheimer's dementia-Control compare in order to understand language deterioration (written example from PPT)
Define evaluation research
Collection and analysis of information related to effects of a program, policy, or procedure - LOOKING AT AND BEING ABLE TO DESCRIBE EFFECTIVENESS BY DESCRIBING WHAT'S GOING ON WITH THE TREATMENT
Data that are described as scores or values that fall under a continuous?
Continuos data
Continuous data
Continuous data so good examples of that is when you collect height. Height runs on a scale 23 more specific 123.6 pops right because it runs across the continuum. So discrete and continuous: so in some questions you're going to be you're going to have to collect on exam questions I tell you before I pass my exam exam was worth 75 points the lowest I'll give you the range the lowest points earning the classes this the highest was this and then what's the one the average the average the mean is this so I collect data on all exams to get an idea of what's the distribution of scores how many students are in how many are above and then how many are in the middle.
Types of exploratory research (3)
Correlational Research Case control studies Cohort studies
What is another name for descriptive research design?
Correlational/observational studies
Discrete data
Discrete data, if you remember, are data that are described as having a finite number of possible values and the space between values on the number line the values must be a whole number. Example: Questions on the test, that's a whole number and then you could determine the number of correct, or the number of vehicles that are owned in the US household to be to 134 but it can't be 2.5 you can't be 3.8 is that that kind of data. Vehicles is a unit of whole
Is prediction causation?
Does NOT mean causation Implies prediction NOT causation
What needs to be considered when using a randomized control trial?
Ethical dilemma of withholding treatment
In what category of research does a meta analysis belong to?
Experimental
What is the most powerful research design? What (5) types fall under this design?
Experimental design 5 types: Randomized controlled Quasi-Experimental Sequential Clinical Trails Single-Subject Design Meta-analysis
4 types or categories of research
Experimental, descriptive, survey, exploratory
4 categories of research:
Experimental: Randomized Controlled, Quasi-Experimental, Sequential, Single Subject Design & Meta-Analysis Studies Descriptive: Cross sectional, longitudinal, Semi-Longitudinal, Historical, Case, Secondary Analysis & Evaluation Studies Exploratory: Cohort, Case Control & Correlational Studies Survey (Can be a category of research but can also be an instrument and can cut across experimental, exploratory and descriptive.)
What category of research does a cohort study to?
Exploratory
Exploratory research example
Home literacy practices and how home literacy practices impact influence children's language and literacy skills development so maybe there's a lot of stuff just on home literacy and maybe there's a lot of stuff on children's language development but there's not a lot of studies bringing these two things together and you want to explore that area more you want to explore the relationship between those two things and be able to explore and come up with you and kind of move the fuel forward and contribute so this type of research is really good and it's very helpful when we're trying to formulate research questions or Define a problem
Example of Correlational Research
How GRE scores are RELATED to graduate level GPA
example of evaluation research
How an agency is properly implementing medicare/Medicaid policies for students who are receiving SLP clinical services
Parametric and non-parametric statistics (These are inferential statistical measures):
Inferential Statistics- Statistics used to make inferences about whether relationships observed in a sample are likely to occur in the larger population (generalizability). Used to compare differences between groups. Parametric: 1.Distribution bound 2. Interval or ratio data measures 3. Normal distribution (equal variance) because we make assumptions about the distribution 4. Large sample 5. These are powerful statistics because they are more sensitive to differences and relationships. Usually prefered method to use 6. Then we make inferences about the population parameters based on a representative sample. Inferences are only valid if the assumptions are met. Non-Parametric: Distribution Free Doesn't compare results to a normal distribution
Define case study
Intensive study of the background, current status, or environmental interactions of an individual, group, institute, or community
Simple Hypothesis
Involves 1 independent variable and 1 dependent variable you're looking at. Example: Looking at one treatment and how that treatment affects children with voice disorders. Then you have to figure out, what am I looking at? Treatment effectiveness? Relationship? Describing the variables that leads then to your design?
When is an ABAB Design used?
It can be used to rigorously test the success of an intervention on a person, school, or community and provide evaluative measures as to the general effectiveness of that intervention.
Weaknesses of case study
Limits to generalizability Imposition of bias
Complex Hypothesis
Looking at more than one variable, multiple variables. Example: Looking at one population, but more than one variable that affects that population.
Hypothesis testing
Making an objective decision about the hypothesis that you made at the beginning of your paper.
The numerical average or scores or values a particular variable is called _______?
Mean
Measures of central tendency (statistics) - mode, mean, median:
Measures of central tendency is being able to compute the mean, the median, and the mode of your data set and this can be done very easily on a computer. You can plug in the numbers on the result but you are the ones that need to interpret what these things mean. So these are measures of central tendency or statistics that indicate average or typical value. So what the Mean is the numerical average of the scores for a particular variable. That's calculated by adding up all the values of all your data points and then is divided by the number of data points. So how do I do that here on the test scores? I take everybody's scores and I add them up and then I divided by the number of students who took the test right. The Median is the numerical midpoint of the scores or values that is at the center of the distribution of the scores.The Mode is the most frequently occurring score for a particular variable.
meta-analysis
Meta analysis always provides an estimate of Effect Size, which involves a Statistical technique that generates an Effect Size. Typically, if the effect size is large, the effectiveness of the treatment is high. If the effect size is small, the effectiveness of the treatment is low.
The inclusionary and exclusionary criteria information do a lit review paper is included in what selection of the paper?
Methods
Is an answer in research ever final? Why?
No answer in research is ever really final, the answers either support something, either prove or disprove something but not with 100% certainty.
Null Hypothesis
No statistically significant relationship between the variables. More of a statistical hypothesis that is tested by running an analysis on a software like, SPSS. NEVER TRY TO PROVE A NULL HYPOTHESIS, ONLY DISPROVE. (We try to disprove that there is no significance by saying that there is a significance) Statistical in nature.
research design has two (threats)
Number one, be aware of it and number two, try to control it under certain conditions in the experiment.
Define correlation research
Predictive research which is designed to predict a behavior or response based on the relationship between the behavior and other variables
Which research design used most often to study something rare?
Quantitative Data
What are the components of a quasi-experimental study?
Quasi = fake Only comparing to self and using as control group No control group of randomization
What does a research plan typically include?
Research question(s) Methods of observation Methods of measurement Conditions Procedures for collecting data Method of data analysis
Example of single subject design
Stuttering using no method, treatment, no method, treatment
Define reliability
The consistency of the measurements. To what level will the instrument produce the same results under the same conditions every time it is used. Adds to the trustworthiness of the results because it is a testament to the methodology, if the results are reproducible. The reliability is often examined by using a test and retest method where the measurements are taken twice at two different times. The reliability is critical for being able to reproduce the results, however, the validity must be confirmed first to ensure that the measurements are accurate.
68-95-99.7 rule
The rule that describes sort of these percentages - the 68th which was in the middle - the 95 which is below or above the mean - Then the 99.7% which falls within the three standard deviation important to understand statistically in order to interpret results
different parts of a lit review
Title Page: Running head, Title of Paper, Author, Institutional Affiliation Abstract: Brief, concise, but comprehensive, Single paragraph, word limit 150-250, keywords Introduction: State problem/Research questions, explore importance/rationale, Historical/ theoretical background Method: Definition of conceptual and operational terms, description of sample of studies, description of selection criteria Results: Statistics and data analysis across sample of studie selected/included, reliability, validity, fidelity Discussion: Evaluate and interpret the implications of the results for the purpose of answering the research question, draw inferences and conclusions, emphasize theoretical/practical consequences of the result, state limitations, discuss generalizability of findings References: accurate & complete in APA format, hanging indent
What is the purpose of a literature review?
To analyze the progress of research towards clarifying a problem
Provide a definition for simple hypothesis
To test one independent variable and test one dependent variable
Working Hypothesis
Used when there is very little research in the area being investigated. They try to answer the question, "Is there a difference?" because there is so little information out there, this hypothesis is guiding the design of the research.
What is meant by person first language?
Value person independent of disability (person first, disorder second)
A fitness instructor wants to test the effectiveness of a performance-enhancing herbal supplement on students in her exercise class. To create experimental groups that are similar at the beginning of the study, the students are assigned into two groups at random (they can not choose which group they are in). Students in both groups are given a pill to take every day, but they do not know whether the pill is a placebo (sugar pill) or the herbal supplement. The instructor gives Group A the herbal supplement and Group B receives the placebo (sugar pill). The students' fitness level is compared before and after six weeks of consuming the supplement or the sugar pill. No differences in performance ability were found between the two groups suggesting that the herbal supplement was not effective. What makes the study experimental? What type of info is collected? Can we establish cause and effect? Are there outside variables that can impact?
What makes the study experimental? - random selection - students are assigned into two groups at random (they can not choose which group they are in) - manipulation - fitness instructor probably needed to be trained - control group - group b, receives the placebo (sugar pill) What type of information is collected? - fitness level is compared before and after 6 weeks of consuming the supplement of the sugar pill can we establish cause-and-effect? - Yes, because of the elements of research (random selection, control group, manipulation) so we are more confident that this is a causal relationship are there outside variables that the participants maybe can do that might have an impact on the results for the outcomes? - students can be training on their own outside of the study - students could be taking other supplements
An investigator wants to evaluate whether a new technique to teach math to elementary school students is more effective than the standard teaching method. Using an experimental design, the investigator divides the class randomly (by chance) into two groups and calls them "Group A" and "Group B." The students cannot choose their own group. The random assignment process results in two groups that should share equal characteristics at the beginning of the experiment. In Group A, the teacher uses a new teaching method to teach the math lesson. In Group B, the teacher uses a standard teaching method to teach the math lesson. The investigator compares test scores at the end of the semester to evaluate the success of the new teaching method compared to the standard teaching method. At the end of the study, the results indicated that the students in the new teaching method group scored significantly higher on their final exam than the students in the standard teaching group. What makes the study experimental? What type of info is collected? Can we establish cause and effect? Are there outside variables that can impact?
What makes the study experimental? - There's a control group (group B) where teacher uses a standard teaching method - random assignment - kids were randomly assigned so everyone had an equal chance of either group A or B - manipulation - teacher probably needed to be trained & selection of the teachers - controlling - inferring that the teacher was trained What type of information is collected? - test scores because it compared test scores at the end of the semester can we establish cause-and-effect? Yes, because of the elements of research (random selection, control group, manipulation) so we are more confident that this is a causal relationship are there outside variables that the participants maybe can do that might have an impact on the results for the outcomes? - students are allowed to study together or talk about different methods they are taught
Why is experimental research the most robust?
by manipulating and controlling the variables then we are more confident in our results we are more confident in the outcomes than we can say yes because we manipulated the independent variable the result of the dependent variable is due to the independent right it's not due to some other variable that we did not manipulate and control
Explain two (2) key elements of an experimental research design.
elements of manipulation and control are really considered the most important in this kind of design is because by manipulating and controlling the variables then we are more confident in our results we are more confident in the outcomes than we can say yes because we manipulated the independent variable the result of the dependent variable is due to the independent right it's not due to some other variable that we did not manipulate and control
Quantitative research characteristics
examine the relationship between variables with the goal being to analyze and represent that relationship mathematically through statistical analysis most commonly used in scientific research problems objective data that are collected and analyzed numerically focus on numbers, measurement, deductive logic, control, manipulation, experimentation
Define exploratory research
exploratory research looks at how events relate to other factors and also looks at relationships
Levels of measurement:
level of measurement classifies the event into a category. A category has to be mutually exclusive. In a nominal way, you're collecting data that are exclusive of each other. Nominal scales. I have here just put people, events, or objects or attributes into categories so that you can count the job. So it puts people or events, perception, objects, or attributes you can be an exclusive category and then you can count the numbers in those categories. Nominal scale is the lowest form of measurement that researchers use because it's simply categorizing the information, anything else about the categories inside the number. Ordinal is. Ordinal ranks your data it's not just about collecting the data but now it ranks it either from highest to lowest or from you know tallest to shortest one step further and then it writes it a little bit on a continuous scale .The disadvantage of using an ordinal scale is that it does not provide information regarding the magnitude of the difference between data points. intervals there is considered no meaningful zero in the scale of measurement, your operations are numerical it does not allow for comparison of absolute magnitude because there's no true zero. Examples of data that you can collect that run on the scale like temperature and IQ Score. Ratio is the highest level of measurement. If you want to know the order of things then you would use an Ordinal, Interval and Ratio. So it really is about the strength of your conclusions about accuracy of the results. Did the results actually measure what was intended to be measured and in my study to measure IQ, did I really collect data that measured IQ? Did I use the right instruments that I can answer my research questions?
Define cohort study
looking at a group they don't have the condition that's under study but you're going to follow them so it kind of a kind of feels like a longitudinal but it's called cohort because it's the group of people under study now
What is the biggest challenge/problem with RCTs?
moral or ethical dilemma about one who's going to get a treatment and the other group is not going to get that treatment so there's although these aren't really really strong studies there is many times it's been identified that. especially in the medical field there are ethical issues to consider in delivering treatment or in withholding treatment you know cancer patients that get into an experimental clinical trial that might have you know my promise them longevity living three more years and then having another group or they're not going to get them and maybe they're going to die right lot of ethical moral issues to consider so here's kind of the visual so I take it I have
Define single subject design
not considered case studies the subject acts as their own control and there are different ways on how the data is collected (ABA design)
What is correlation?
statistical method for measuring relationship/(or association) between two or more variables different types: is a correlation is not a correlation. A positive or negative. (correlation is NOT causation)Seeks to describe a relationship between two variables, such strong, negative, positive, weak or statistically significant. The Higher the value, the stronger the relationship. .10-.30= Small to Low relationship .40-.60= Moderate relationship .70+=High relationship
purpose/goal of quantitative research
test theory establish facts and validate show relationships explain and predict statistically describe
Construct validity
the correct ability to measure if you have high khans from validity on a test Please that every test item is measuring what it says it's going to measure but if you have low cost from validity and means that the items are not testing the construct the ability to skill that you want to measure validity is each item is Judge for its applicability to the characteristics.
Exploratory research sets...
the priorities and the research agenda for future in a field it also is a good type of design to develop hypotheses so it helps to formulate research questions or define problems it allows us to isolate key variables to look at to examine the relationships it establishes Priority for future research and it can help to develop hypotheses so what kind of studies are we looking at
External validity
tries to answer the question to what extent can the results of a study be generalized. External validity of is all about now that I have my results can I generalize it to the larger population? and that's important because that's why we do research how else could we get our results?
Exploratory research is often conducted when....
you want to explore a topic or you want to do more searching on a particular topic or you want to clarify or Define a problem that has been posed in the literature or you have a situation where you just need more insight
