Exam 2- ch. 4, 5-9, & 11

Summarize the description, quantitative counterpart, and methods of the following criteria that are used to establish trustworthiness of qualitative research: confirmability

the extent to which findings can be corroborated by others. efforts must be taken to ensure that the researchers perspectives do not distort the voice of the participants. reflexivity, or the process of identifying a researchers personal biases and perspectives so that they can be set aside is necessary to make the views of the researcher transparent

Summarize the description, quantitative counterpart, and methods of the following criteria that are used to establish trustworthiness of qualitative research: dependability

the extent to which qualitative data are consistent. one way consistency can be examined is in terms of time. multiple time points are preferable to data collected at only one time point. another way to examine consistency is across multiple coders. when coding transcripts, two or more researchers can code independently and then compare their results.

Summarize the description, quantitative counterpart, and methods of the following criteria that are used to establish trustworthiness of qualitative research: transferability

the information from a qualitative study may be extended, or applied to other situations the burden lies primarily on the practitioner who is interested in applying the results

Define the following terms and how they influence qualitative research: constructivism

the philosophy that our understanding of the world is "constructed", reality is filtered by our experiences and our past knowledge, always subjective and interpreted the constructions that exist to explain our world are inseparable from the people who give those constructions meaning

Describe the key characteristics of the following data collection methods: participant observation

the researcher engages with the participants in their naturally occurring activities in order to gain more in-depth appreciation of the situation. a more immersive form of data collection

Describe the key characteristics of the following data collection methods: field notes

the researcher is removed from the experience and observes/reports from afar the most unobtrusive way of recording and observation. the focus is on watching and listening. field notes often describe both what is seen and the researcher's impressions.

Definition of prolonged engagement

the researcher spending enough time getting to know individuals that a sense of trust and familiarity is established

Describe the key characteristics of the following data collection methods: open-ended interviews

there are no set questions, so the process directs the questioning. typically done face-to-face so the interviewer is free to probe, ask follow up questions, and take the lead of the interviewee to pursue new and possibly unexpected areas of inquiry

characteristics of prospective cohort studies

Look forward in time- Research question is identified before the study begins, then individuals are followed over time to see who does and who does not develop the condition or outcome.

from her voicethread about difference in means what does difference in means, mean?

Many rehabilitation studies are interested in differences. intervention studies seek to determine whether there is a difference between the intervention and control groups, or whether there are differences before and after treatment. a descriptive study may examine the differences between people with and without a condition.

Definition of audit trail

collection of documents from a qualitative study that can be used to confirm the data analysis of the researcher (promotes confirmability by making the data available to outside sources)

review: what are the two primary approaches to statistical inference?

confidence intervals and hypothesis testing

Recognize the key purposes, characteristics, and examples of the following research designs: ethnography

describe a group of people, their behaviors, and/or their culture. Insider point of view is used to understand the larger culture or social structure Ethnographic researchers become immersed within the culture being studied. The ethnographer observes and, once invited, participates in the routines, rituals, and activities of the group, which is referred to as participant observation. unlike phenomenology, which describes, ethnography explains.

Definition of thick description

enough details are provided about the people, situations, and settings that readers can determine the transferability to their own situations

what is the main purpose of a t-test

examine the differences in means Look at picture I took from her voicethread

Definition of correlation

in a simple correlation, the association between two variables is determined. correlational studies are cross-sectional, with data collected at a single point in time. at least two measures are administered and related, however multiple measures can be administered

Definition of triangulation

multiple resources and methods are employed to verify and corroborate data; several methods lead to the same result in each case the inclusion of multiple participants and observers is one way to accomplish triangulation. can also be achieved by collecting data using different methods, such as interviews, focus groups, and participants observation.

Compare and contrast definitions of prolonged engagement, triangulation, member checking, thick description, and audit trail

next slides

Describe the key characteristics of the following data collection methods: artifacts

objects that provide information about the subject of interest for example, in a study exploring factors that facilitate return to work for individuals with spinal cord injury, a study asked to see assistive technology that was used in the workplace. actual photographs of these artifacts are included in the study report.

Definition of member checking

participants are regularly and repeatedly queried to ensure that the researcher's impressions are accurate for example, during initial interviews, the researcher asks follow-up questions and/or repeats statements back to the participant.

What is client-centered practice and what is the central construct of this approach according to Hammell (2013)?

Client-centered practice is an approach to practice that attaches importance to the individuals values and preferences, and respects the expertise that the person brings to the situation in the form of lived experience. Hammel (2013) claims that the central construct of client-centered practice is respect, which includes respect for the clients and their experiences, knowledge, and right to make choices about their own lives. Clients want to be listened to, cared for, and valued. Client-centered practice can enhance therapy outcomes because clients must engage in the process in order to benefit from therapy.

What is a decision aid

Decision aids are materials that provide information to the client to support the decision-making process. They may take the form of written materials, videos, interactive internet presentations, or other resources.

Define purposive sampling and snowball sampling and recognize examples of each.

-purposive sampling: participants included in the study and the setting it takes place are selected for a purpose or a specific reason -Snowball sampling- initial participants are asked to recruit additional participants from their own social networks. useful because carefully selected initial participants may be able to reach and recruit individuals who the researcher does not have a way to identify or contact

recognize research questions that are consistent with the qualitative research tradition.

-theory is not developed, hypotheses are avoided -broad and general questions instead of specific -stay away from words such as cause or relate; use words such as discover, inquire, describe, and explore -being with "what" or "how" instead of "why"

definition of Cohen's d

A common effect size statistic that is easy to understand. It measures the difference between two group means reported in standard deviation units. .2 to .5 is a small effect, .5 to .8 is a medium effect, and anything greater than .8 is a large effect. Because Cohen's D is expressed in terms of standard deviations, the d value can be greater than 1.

her notes: what is the difference between embedding data and merging data? how would you tell an example of each apart? -embedding data-

A researcher who uses a mixed method design that includes both quantitative and qualitative data needs to make a decision about the approach used for the data. Embedding data means the research will select one type of data (quantitative or qualitative) as the primary focus in a study and supplement with another type of data. For example, a researcher who wants to understand quality of life for individuals with disability might summarize findings for a quantitative study using the SF-36, a measure of quality of life. Then they might add to the findings by making a statement something like....These findings are supported (or different) from a short answer survey of individuals with disability in which one of the questions was "How would you describe your quality of life?"

her notes: what is the difference between embedding data and merging data? how would you tell an example of each apart? -merging data-

A researcher who uses a mixed method design that includes both quantitative and qualitative data needs to make a decision about the approach used for the data. Merging data means the researcher will use both quantitative and qualitative data (roughly equally) to summarize the results. For example, a researcher who wants to understand quality of life for individuals with disability might summarize findings using: 1) the SF-36 (a quantitative measure of quality of life, and 2) the findings from a qualitative interview of individuals with disability in which one of the questions was "How would you describe your quality of life?

her notes: can you provide examples to differentiate between phenomenology and narrative designs?

Both of these approaches commonly use interview to gather qualitative data. But the purpose of each is slightly different. In phenomenology, the purpose is to understand the 'lived experience' so it has a narrower focus than narrative. An example of a phenomenology study purpose might be to understand the daily experience of using public transportation after having a spinal cord injury. In narrative, the purpose is to gather stories about many aspects of life, often in chronological order. An example of a narrative study might be to gather stories of the protests that led to the signing of the ADA.

What are the characteristics of the consensus building and agreement stages of shared decision making and how might these stages vary depending on the client?

During consensus building, the therapist and client share relevant information. They discuss client values and preferences, important information about the clients meaningful activities, then the therapist shares information about intervention options, pragmatic issues, and evidence to support interventions. During the agreement stage, shared decision-making results in an agreement about the direction that the intervention will take.

What is the main outcome of data analysis in qualitative research?

Identified patterns within data are labeled or described as themes. Themes are what the reader sees. the results section of a qualitative study reports the themes, describes what the themes mean, and illustrates the themes through actual quotations from the study participants

What is the progression of communication in a shared decision making approach?

Listen, speak, and then listen again. The therapist listens to the client regarding the client's values, preferences, and particular life circumstances. In the context of the information provided by the client, the therapist presents options and explains the pros and cons of each. The therapist then listens to the client as he or she weighs the options. The therapist and client continue the discussion, which is focused on ultimately making a treatment decision.

Who are the other people that may be involved in shared decision making?

The inclusion of the therapist and client is the obvious component. From the clinical side, therapists typically work on a team, and other team members may be invested in the treatment decision. On the client's side, there may be family members or other individuals who are concerned about the treatment decision. It is especially important to involve other individuals who are close to the client when they assume caregiving responsibilities and will be directly involved in carrying out the intervention, and when they are impacted by the intervention choice.

What strategies might an occupational therapist use to engage a client in shared decision making?

The physical environment in which the discussions take place is important. A space that is free from distraction and allows for private discussion will facilitate greater connection between the therapist and client. The manner in which the therapist communicates can go a long way toward promoting inclusion of the client in shared decision-making. The therapist's nonverbal cues, such as an open and forward-leaning posture, eye contact, and head nods, can indicate warmth and attentiveness. Open-ended questions that promote discussion are more useful than questions that can be answered with a single-word response, and active listening with paraphrasing will likely encourage more openness on the part of the client. Once the therapist educates the client about shared-decision making, he or she can determine the client's desired level of involvement.

What are the contributions of the professional and the clients in shared decision making?

The professional brings information about the clinical condition and options for intervention, including the risks and benefits of each option. Clients bring their own information, which includes values, preferences, lifestyle, and knowledge of their situation. The therapist and client then work collaboratively to arrive at clinical decisions.

Recognize the key purposes, characteristics, and examples of the following research designs: grounded theory

The purpose is to develop new theory from the data collected. Begins without a hypothesis or assumption. Data is collected concerning a general question or topic. The theory comes out of the data, or, in other words, is "grounded" in the data.

What type of information is typically included in a decision aid

They usually include: -An explanation of the condition to help the client understand their condition and how interventions can work to target specific aspects of it -Identification of the decision that needs to be made, including the different interventions that are available -Options and potential outcomes based on scientific evidence. Summary of evidence for each option. -Questions to help clients clarify their values. This is often written in a workbook format with space for the clients to write their responses. It could focus on activities the individual wants to return to, desired outcomes, troublesome aspects of the condition, living arrangements, social support, financial resources, insurance coverage, and so on.

what is a logistic regression

a statistical method for analyzing a dataset in which the outcome is measured with a dichotomous variable (i.e. there are only two possible outcomes)

Describe the key characteristics of the following data collection methods: focus groups

allows multiple individuals to be interviewed at once. in focus groups, an interview is conducted with a group of individuals to target a particular topic. although it may be more difficult to get in-depth information from each individual, the dynamic nature of a focus group provides participants with the opportunity to bounce ideas and thoughts off one another. the group members may confirm an individuals experience or provide a different point of view.

Recognize the key purposes, characteristics, and examples of the following research designs: mixed methods

both quantitative and qualitative methods are used to increase the breadth and depth of understanding a research problem. Can be done by merging the two together, connecting one set of data to a second set, and embedding a secondary supplement into a primary source.

Recognize the key purposes, characteristics, and examples of the following research designs: narrative

can be characterized as storytelling. storytelling involves remembrances, retrospectives, and constructions that may focus on the recounting of an event or series of events, often in chronological order. many narratives feature epiphanies or turning points in an individuals life

Recognize the key purposes, characteristics, and examples of the following research designs: phenomenology

understand and describe the lived experience from the point of view of the research participant. Useful in situations that are poorly defined or potentially misunderstood. Emphasis on description and insider perspective. typically one person or a small, selected group of participants are included in the study. primary methods of data collection include in-depth interviews, discussion, and observation.

Summarize the description, quantitative counterpart, and methods of the following criteria that are used to establish trustworthiness of qualitative research: credibility

when it accurately reflects the reality of the research participants one measure lies in the sample selection requires that the researcher use methods to ensure the participants respond honestly and openly triangulation is another strategy that can enhance the credibility of a study- multiple resources and methods are employed to verify and corroborate data, the use of several methods leads to the same results in each case

What is a response rate in survey research and why is it important to note?

-% of individuals who return a survey based on the total numbers of surveys administered -If large numbers of individuals choose not to respond, it is likely that there is a response bias

range of values for correlation coefficient

-1 (perfectly inverse) to 0 (none) to +1 (perfect positive) .80-1.0 strong .60-.80 moderately strong .40-.60 moderate .20-.40 moderately weak 0-.20 weak

recognize the notation, key characteristics, and examples of: a. designs without a control group

-A design without a control group is technically not considered an experimental design. This is because an experimental design technically has to have a comparison between two groups. But, a pretest-posttest design is important because it is used often in literature. -Researchers typically use this design in order to decide if an intervention has the potential to even make a difference before spending a ton of time and money on doing an extensive randomized control trial. -Designs without a control group are sometimes referred to as pre-experimental designs because the intent of them is to examine a cause and effect relationship. But, like I mentioned before, they don't have a control group so it limit's researchers' abilities to conclude a relationship. -Designs without a control group have a lower level of evidence (IV) which means it is weaker than a randomized control trial and non randomized control trials. -The notation for this design is O X O

What are some characteristics that should be considered when evaluating descriptive and predictive studies?

-Cannot be analyzed using hierarchical levels because there is lack of random assignment to the groups and no manipulation of the independent variables -Evaluating predictive and descriptive studies --Control for alternative considerations --Matching - improves the strength of a nonrandomized design --Sample size --Sampling bias --Measurement methods should be evaluated

Know the range of values that are possible for a correlation coefficient and summarize the strength, direction, and significance.

-Correlations range in strength from 0 to 1.0 and are reported as r values. Values can range from -1 to 1. -The direction of a correlation can either be positive or negative. A positive correlation means that two variables are related in the same direction; as one variable increases, so does the other variable. For example, speed and distance are generally positively correlated, meaning that the faster you are, the more distance you can cover. In contrast, speed and time are negatively correlated; the faster you are, the less time it takes to get to a destination. -Strength: the extreme values of -1 and 1 indicate a perfectly linear relationship where a change in one variable has a perfect change in the other. A coefficient of zero means there is no relationship. When the value is inbetween 0 and -1/1, there is a relationship, but the points do not fall all perfectly on a line. As r approaches -1 or 1, the strength of the relationship increases and the points fall closer to a line. -Direction: positive coefficients indicate that when the value of one variable increases so does the other variable. This makes an upward slope on a scatterplot. Negative coefficients represent when the value of one variable increases, the value of the other variable decreases. This makes a downward slope.

What is the difference between descriptive and inferential statistics?

-Descriptive statistics describe the data in a study. They provide an analysis of data that helps describe, show, or summarize it in a meaningful way such that patterns could emerge from the data. -Inferential statistics are techniques that allow us to use study samples to make generalizations that apply to the population. -Descriptive statistics are used in the calculation of inferential statistics. -An important element of inferential statistics is "infer"- used when a researcher wants to infer something about a larger population based on the sample used in the study. They are often divided into two categories: 1- test of differences (e.g. t-tests and analysis of variance) and 2- tests of relationships (e.g. correlations and regressions). With inferential statistics, a test is conducted to determine if the difference or relationship is statistically significant.

Less wordy answer to: What is the difference between descriptive and inferential statistics?

-Descriptive stats describe the data in a study (an analysis of data that helps us describe, show or summarize it in a meaningful way) -Inferential stats are techniques that allow us to use study samples to make generalizations that apply to the population o Descriptive stats are used in the calculation of inferential statistics

recognize the notation, key characteristics, and examples of: a. non-randomized controlled trials

-Design same as randomized control trial, except for group assignment --Control/intervention groups --Not an equal chance of being in each group --Lower level of evidence than RCT -Best for studies where randomization is not possible --Uncontrollable differences (age/race/sex/gender) --Pre-existing groups (school/activity involvement/college major) -Cluster randomized control trial --Condition randomly assigned to setting --Group assignment based on setting, still not true RCT notation on page 113

Recognize definitions, examples, and statistics of inter-rater reliability

-Different people give the same test and get comparable results -Important for accurate and dependable scores -Clear methods and instructions -Determining inter-rater reliability --Same test given to the same client at the same point in time, but administered by different people --Cohen's kappa: used to assess inter-rater reliability with categorical or qualitative measures

Recognize definitions, examples, and statistics of internal consistency

-How well is a test measuring what you want it to measure? -For a measure with multiple items, questions, or subscales, internal consistency refers to the unity or similarity of items -Each item in a specific test is hopefully measuring the same thing -Important when all the items of a measure are expected to assess the same construct --Example: a test specifically measuring for depression should only contain items measuring this specific construct, not items that would be more representative of anxiety --Measurement: Cronbach's alpha

Compare and contrast definitions of incidence and prevalence, including the basic statistical formulas.

-Incidence- Frequency of new occurrences of a condition during a specific time period; # of new cases --Equation: = number of new cases during a period of time / total population at risk -Prevalence- number of individuals in a population who have a specific condition at a given point in time; how widespread a condition is --Equation: = number of cases at a given time point / total population at risk

Compare the levels of evidence in prognostic studies versus efficacy studies (Table 1-3).

-Level I for efficacy studies: systematic review of randomized controlled trails -Level I or prognostic studies: systematic reviews of prospective cohort studies -Level II for efficacy studies: randomized control trial -Level II for prognostic studies: individual prospective cohort study -level III for efficacy studies: nonrandomized control trial -Level III for prognostic studies: retrospective cohort study -Level IV for efficacy studies: one group trial with pretest and posttest -Level IV for prognostic studies: case control design Level V for efficacy studies: case reports and expert opinion Level V for prognostic studies: expert opinion, case study

What is validity and what are the characteristics of a valid assessment?

-Measurement validity: ability of a test to measure what the test is intended to measure Characteristics: ?

Compare and contrast norm-referenced and criterion-referenced measures and recognize examples of each.

-Norm-referenced measures: a client's scores are compared with those of other individuals --Used to determine if an individual's abilities fall within the typical range --Ex. IQ Test -Criterion-referenced measures: a test based on standard or fixed point, which is established by experts --Used to determine how an individual's performance compares to the bench mark --Ex. Functional Independence Measure

Characteristics of Case Control Designs

-Observational, retrospective, cross-sectional study that can be used to answer prognostic research questions concerning which risk factors predict a condition. -Commonly used in epidemiological research that is conducted after a condition has developed. -Individuals who already have a condition constitute one group - they are matched and compared with individuals without the condition. -No random assignment. -Answer prognostic and predictive questions and may use analyses that are more suitable for answering predictive questions

recognize the notation, key characteristics, and examples of: a. crossover trials

-Participants are randomly assigned -Level II evidence-Same level of evidence in the hierarchy as RCTame level of evidence -In crossover designs, participants all get the same treatment, but in a different order, for example one group gets intervention A while the other group gets intervention B, and then they switch -Mostly used when permanent change is not expected -This is to avoid a history threat, because if a permanent change was expected, the second intervention used would be affected by the first intervention used. --Common in assistive device/technology intervention studies -Example: A study comparing scheduling tools in individuals with TBI-traditional paper calendar compared to a phone application with reminders notation on page 110

recognize the notation, key characteristics, and examples of: a. randomized controlled trials

-Randomized Controlled Trials include at least two groups (typically an experimental group and a control group), and the participants are randomly assigned. -useful for single studies and is rated as Level 2 evidence. -Randomized control trials with large sample sizes that target the intervention support the conclusion that the intervention caused the positive outcome. -notations on next notecard...

Define the two main measures of variability?

-Range and standard deviation are the two main measures of variability. -Variability refers to the spread of scores in a distribution. Distributions with the same central tendencies can still be very different because of the variability in the scores. Measures of variability is range and standard deviation. -Range is one measure of central tendency that indicates the lowest and highest scores. For example, the age range of participants could be 18-49. -The most common measure of variability is standard deviation. Standard deviation is the expression of the amount of spread in the frequency distribution and the average amount of deviation by which each individual score varies from the mean. A large standard deviation means there is a high degree of variability and a small standard deviation means that there is a low degree of variability.

What is reliability and what are the characteristics of a reliable assessment?

-Reliability describes the stability of a test score. a reliable assessment measure is one for which the scores are expected to be trustworthy and consistent. -Assessment measures are expected to be dependable and have reduced amount of measurement error -Characteristics: more questions on test; standardized- specific procedures to follow for environment, structure, and scoring

Recognize definitions, examples, and statistics of test-retest reliability

-Test-retest reliability: same test is given to the same client at two different points in time - how similar are the scores? -Stability, repeatability of a test over time, consistency of scores -Determining test-retest reliability --Time is a big factor --Enough time in between so memory or practice are not factors --Not too much time that history or maturation are factors

What is the PEDro scale and how is it used in EBP?

-The PEDro scale is a numerical rating applied to individual studies to objectively assess the methodological/study design quality. It's an 11-item scale that rates external validity, internal validity, and reporting of outcomes. 1: eligibility criteria 2: random allocation 3: concealed allocation 4: baseline comparability 5: blind subjects 6: blind therapists 7: blind assessors 8: adequate follow-up 9: intention-to-treat analysis 10: between-group comparisons 11: point estimates and variability

What is a standardized test and what are specific aspects of the test that may be standardized?

-The administration of a test is the same for all clients taking the test -Specific procedures are followed for the environment, instructions, and the scoring -Enhances reliability by reducing the differences in environment for individuals completing the testing -All procedures must be followed to be valid

Define the three main measures of central tendency

-The measure of central tendency describes the location of the center of a distribution. The three measures of central tendency are mode, median, and mean. -The mode is the score value that occurs most frequently in the distribution. The mode provides information about the distribution, but it is generally of less use than the other measures of central tendency. Greatly influenced by chance. -Median is the score value that divides the distribution into the lower and upper halves of the scores. This is most useful when distributions are skewed, because it is less sensitive to extreme scores. -Mean is the same as the average and balances the scores above and below it. Calculated by summing the scores and dividing the sum by the number of participants. Used most often in research, particularly when calculating inferential statistics.

recognize the notation, key characteristics, and examples of: a. retrospective intervention studies

-The researcher looks back at something that has already occurred and uses existing records to collect the data. -retrospective studies are not experimental because the independent variable is not manipulated. instead, they are observational. Sometimes are called retrospective cohort studies because they utilize and compare existing groups (cohorts) of individuals.

notations for randomized control trial

-There are a few different notations for randomized control trials. -The notation for a "true" randomized trial is ---R O O (control group) ---R O X O (experimental group) ---This type of no-treatment control group is usually avoided in therapy studies for ethical reasons. -Because no treatment controls are avoided, here is the notation for comparing a new intervention with usual treatment or standard therapy ---R O Xa O (usual treatment) ---R O Xb O (new intervention) -Another research approach is comparing a standard intervention with a standard intervention plus a new intervention. Here is the notation for the combined treatment. ---R O Xs O (standard intervention) ---R O Xs + Xa O (standard intervention plus a new intervention) (see page 108)

How does sample size affect the conclusions that are made from intervention research?

-a larger sample reduces the likelihood of making a type II error (i.e. when the researcher finds no difference between groups, but actually a difference exists) and in general reduces sampling error so that the results of the study are more likely to reflect the true population.

what are the two primary approaches that are used for statistical inference?

-confidence intervals: an approach to estimate a population parameter from a sample statistic -hypothesis testing: an approach to test an assumption about a population parameter from a sample statistic

Compare and contrast continuous (quantitative) and discrete (categorical) data.

-continuous data result from a test in which the score can be any value within a particular continuum. for example, range of motion is continuous data expressed in terms of the number of degrees of movement within a 360 degree range -Discrete data (categorical), are obtained when classifying individuals or their performance into groups, such as gender or diagnosis. Can be numerical, although the numbers assigned reflect a category more than a quantity. -Dichotomous: a type of discrete data with only two categories: typically something exists or does not exist. for example, a child does or does not have autism.

Why is cost-effectiveness an important outcome in intervention research and what is a common measure of cost-effectiveness?

-cost effectiveness is important because policy makers want to apply resources sensibly by spending money on things that influence health the most; likewise, clients want to use their health-care dollars wisely -cost effectiveness studies often use the quality adjusted life year (QALY) to assess the impact of an intervention. QALY combines quality of life assessment with the number of years that are added to the life by the intervention -Quality of life- measured on a scale 1-0 -Calculated by multiplying the number of years of extra life by the quality-of-life-indicator (For example, if an intervention extends a person's life by 5 years and their quality of life is a .7 on the scale of 1-0, the QALY value is 3.5.)

What are the definitions of descriptive and predictive studies?

-descriptive studies- explains health conditions and gives information on incidence and prevalence of conditions. provides practitioners with info on comorbidities that are common with a particular condition. -predictive studies provide information about factors that are related to a particular outcome. for example, a predictive study might identify what activities of daily living are most important for a successful discharge to home and which motor abilities are most likely to result in independence in mobility.

what statistic is used for a t-test and what statistic is used for an ANOVA test

-for a t-test a t statistic is used -with an ANOVA test, an f-test is used

recognize the notation, key characteristics, and examples of: a. factorial designs

-have more than one independent variable (or factor) -typically used to determine if has differing affected on the additional variable Uses create Varying designs: -2 x 2 design→ 2 levels of intervention and 2 levels of gender a & b=interventions m & f= gender -2 x 3 compares → 2 different interventions (1st ind var) with 3 different settings (2nd ind var) -2 x 2 x 3→ 2 different interventions (1st ind variable) with 3 different settings (2nd ind var.) AND 2 levels gender (3rd in var.) Additional recruitment challenges-larger sample size necessary to allow for additional comparisons (maybe look this up online so its not so confusing)- notation on page 114

What is a Likert scale and what is the type of data in a Likert scale?

-in a likert scale, individuals respond to a range of responses, most typically on a continuum, such as "strongly agree", to "strongly disagree", or "never" to "always". These responses are then assigned numerical ratings. -Controversy over whether to classify into discrete or continuous categories -Currently considered continuous**

Understand the definitions and distinctions in between-group comparison, within-group comparison, and interaction effect and recognize examples of each.

-in between-group comparison: comparison that identifies the differences between two or more groups. These are often the control and intervention groups in most efficacy studies. Can occur either before or after an intervention -within-group comparison:makes a comparison within the same group. Most often, a comparison examines differences in the pretest and posttest scores for each group separately. -interaction effect: Combines the between- and within-group comparisons and is the most important comparison in an intervention study. It is a way to determine whether the intervention group improved more than the control group. It determines whether there was a difference in the way one group performed from pretest to posttest, compared with the other group's performance from pretest to posttest.

What notation is used to summarize the key characteristics of different research designs?

-research design notation is a system that uses characters to diagram the design of intervention studies. -Primary characters include: R=randomization N=nonrandomization X=treatment O=dependent variable or outcome

recognize the notation, key characteristics, and examples of: a. single subject designs

-single-subject designs do not aggregate the scores of participants. instead, the results of each individual are examined separately to determine if the intervention was effective. the basis of a single subject design is to compare an individuals response under different conditions. -each participant is his or her own control -replication is important and occurs in the form of participants; each participant replicates the design.

What are the two categories that inferential statistics are often divided into?

1) tests of differences (e.g., t-tests and analysis of variance) and 2) tests of relationships (e.g. correlations and regressions) With inferential statistics, a test is conducted to determine if the difference or relationship is statistically significant

from the voicethread: what are the 7 steps in hypothesis testing?

1. state the null and research hypothesis 2. select a significance level (alpha) or the probability of making a type 1 error 3. identify the test statistic that will be used to test the null hypothesis 4. describe the key characteristics of the sampling distribution for the test statistic 5. develop a decision rule for when you will reject the null hypothesis 6. using data from the sample, obtain the value of the test statistic and compare to decision rule 7. make a decision regarding the null hypothesis

Recognize the inferential statistics that may be used to analyze differences in means.

?? is this answer right?? There are many statistical tests for comparing differences, but the most common difference statistics are the t-test and analysis of variance (ANOVA). In these tests, the means of the two groups and/or time periods are compared, while taking into account the standard deviations of the means. If a difference is found within the sample, an inference is made that a difference would exist within a larger population. -The t-test is the most basic of inferential difference statistics because it examines the difference between two groups at a single time point, or one group at two time points. -When more than two means are compared, an analysis of variance (ANOVA) is the appropriate test. With a t-test, the t statistic is used. With an ANOVA, an F-test is used to compare means, and the F statistic is used.

Define and give examples of: sensitivity

Ability of a test to detect a condition when it is present, which is also known as true positive. a sensitive test will accurately identify individuals who have a specific condition. however, in doing so, overdiagnosis may occur, such that people are diagnosed with a condition they do not have. this is known as false positive.

her notes: explain odds ratio vs. hazard ratio

Both ratios allow us to compare estimates of 'an occurrence' in one group vs a second group. A ratio by definition is written as A/B=#. An odds ratio has multiple applications and is more general. A hazard ratio focuses on some aspect of risk. The interpretation is the same. A ratio of 1.0 means no difference in the groups. A ratio of > 1.0 means that the A Group is # times more likely to have an occurrence than the B group. A ratio < 1.0 means the A Group is # times less likely to have an occurrence than the B group

Given the formulas for sensitivity and specificity and a small table, calculate the values for sensitivity and specificity and interpret the meaning.

Come back to this when studying but-- -- sensitivity = a/ (a+c) --specificity = d/ (b+d) ex: a test with a 90% sensitivity will correctly return a positive result for 90% of people who have the disease, but will return a false negative result for 10% of the people who have the disease and should have tested positive A test with a 90% specificity will correctly return a negative result for 90% of people who do not have the disease, but will return a positive result, a false positive, for 10% of the people who dont have the disease and should have tested negative

from her notes: concurrent validity

Concurrent validity studies are slightly different in that they look at the relationship between a new measure of a construct and an established measure of a construct.

Recognize definitions, examples, and statistics of concurrent validity

Concurrent validity: used to predict scores on another measure Ex. if you score well on the practice test, you would score well on the exam

Recognize definitions, examples, and statistics of construct validity

Construct validity: ability of a test to measure the construct that it is supposed to measure Ex. an intelligence test being able to accurately measure intelligence rather than education level or memory

from her notes: can you further explain convergent validity?

Convergent validity studies are used to determine if 2 (or more measures) of the same construct are related to each other. For example, a 2004 study by Hotchkiss et al. examined the relationship between three established measures of the construct 'fear of falling'. They found a strong correlation between scores on two of the measures, but the third measure was only moderately correlated with the other two measures. So the third measure must not measure the exact same construct.

Recognize definitions, examples, and statistics of convergent validity

Convergent validity: used to find evidence that the new measure is similar to a measure of the same construct Ex. a measure of self-esteem overlapping with a measure of confidence

from her notes: further explanation of discriminant validity:

Discriminant validity studies are used to determine if a measure of a construct discriminates between two groups of people. For example, let's say a researcher has developed a measure of a proposed construct 'environmental barriers associated with having a disability'. The researcher believes that this measure will help differentiate between the barriers that individuals with and without disabilities experience. To test this belief, the researcher administers the measure to a sample of people with disabilities and a sample of people without disabilities. If there is a significant difference in the scores of the two groups, then the researcher has support that the measure discriminates between the two groups.

Recognize definitions, examples, and statistics of discriminant validity

Discriminant validity: finding a test is able to differentiate between groups of individuals Ex. a test can differentiate between groups that are happy and sad

from her notes: further explanation of divergent validity:

Divergent validity studies are used to determine if two related constructs are actually different from each other. For example, let's say a researcher has developed a measure of the construct 'life satisfaction'. The researcher believes 'life satisfaction' is distinct from related constructs, for example 'quality of life'. To test this belief, the researcher administers the measure of life satisfaction and a measure of quality of life to a sample of people. If the correlation between the two measures is not strong (say, it is moderate of r=.60), then the researcher has support that these are two distinct constructs.

Recognize definitions, examples, and statistics of divergent validity

Divergent validity: finding that the index measure is not related to measures of irrelevant construct Ex. a measure of self-esteem NOT overlapping with irrelevant constructs that may be present in the study such as intelligence

definition of effect size

Effect size describes the magnitude or strength of a statistic (i.e. the magnitude of the difference or the relationship). A way to evaluate the magnitude of a difference. Not all researchers report the effect size but it is becoming more valued as an important statistic.

Compare and contrast the following terms: floor effect and ceiling effect

Floor effect o Test is so difficult, or the construct is so rare that almost all individuals receive the very lowest score Ceiling effect o Test is too easy, and everyone gets a high score to begin with, there is no room for improvement

Be able to interpret a p value at alpha=.05 in terms of 1) the appropriate decision related to the null hypothesis (reject, do not reject) and the decision related to significance (statistically significant, not statistically significant).

If a p-value is less than .05, that means we reject the null hypothesis and it is statistically significant. So we reject the null hypothesis and accept the alternative hypothesis. If the p-value is greater than or equal to .05, we do not reject the null hypothesis and conclude there is no difference or no relationship. Alpha is the probability of rejecting the null hypothesis when it was in fact true. Its value is often 5%. but you can change it to .01 or 1% if you want to be more certain Look back at exam 1 questions for p-value and what type of error each one makes

her notes on multiocollinearity:

In multiple regression, we are interesting in finding the best set of quantitative independent variables for predicting one quantitative dependent variables. For example, for a falls program for frail older adults, I might look at measures of the independent variables: balance, fear of falling, flexibility, speed of walking, etc. My dependent variable might be: number of falls within one year When we are selecting our best set of independent variables we don't want too much overlap or redundancy in the independent variables (or multicollinearity). For example, it wouldn't make sense to use both the Berg Balance Scale and Tinetti Balance Scale because they both measure balance.

Summarize the relationship between reliability and validity.

It is possible for a test to have excellent reliability but poor validity. the test may provide consistent results, but still not test what it is intended to test. however, a test with poor reliability can never have good validity, because reliability affects validity. if a test lacks consistency and stability, ti cannot assess what it is intended to asses

characteristics of retrospective cohort studies

Look back in time- existing records or the client's report on past behavior is used to determine if changes occurred over time.

characteristics of cohort studies

Observational but differ from a case-control design in that participants are followed over time, making this design longitudinal. A risk factor is identified, and then the study follows individuals with and without the risk factor. At a certain time, the risk factor is analyzed to determine its impact on the outcome. Risk ratios are used to interpret the risk of developing one condition if exposed to another, and it takes time into account.

Definition of odds ratios

Odds ratio: Measure of an association between an exposure and an outcome. The odds ratio is an estimate of the odds when the presence or absence of one variable is associated with the presence or absence of another variable. Odds ratio uses a 2 X 2 table and the formula is OR = AD/BC. Odds ratios are interpreted as the odds that a member of a group will have a particular outcome compared with the odds that a member of another group will have that outcome. OR>1 means greater odds of association with the exposure and outcome OR= 1 means there is no association between exposure and outcome OR<1 means there is a lower odds of association between exposure and outcome

from her notes: predictive validity

Predictive validity studies look at whether a measure of a construct (e.g., ACT score) is predictive of a future construct of interest (e.g., college GPA).

Recognize definitions, examples, and statistics of predictive validity

Predictive validity: finding that a test is capable of predicting an expected outcome Ex. how likely it is for a test score to predict future job performance

What are some examples of graphs and figures that are used in descriptive statistics?

Qualitative: •Frequency distribution- when a graph is used to depict the count - Frequency tables •Bar graph •Pie chart •Contingency tables Quantitative: •Histogram •Stemplots •Dotplots •Scatterplots •Boxplot

Discuss why we would summarize an assessment by stating "there is evidence to support the reliability [or validity] of the [name of] assessment" RATHER THAN "the [name of] assessment is reliable [or valid]".

Reliability and validity studies are specific to a certain population or context. For example, let's say a study found good test-retest reliability on a measure of cognition for a large sample of individuals who have had stroke. Can we say that the measure has good test-retest reliability for a sample of individuals who have traumatic brain injury? For example, let's say a study found good construct validity on a measure of environment participation for a large sample of children with disabilities who come from a high socioeconomic background. Can we say that the measure has good construct validity for a sample of children with disabilities who come from a low socioeconomic background?

Define and give examples of: specificity

Specificity is the ability of a test to avoid detecting a condition when it does not exist, otherwise known as true negative. likewise, mistakes can occur with specificity, and some individuals may be missed who do have the condition, resulting in a false negative. look at figure 7-3 on page 137

compare and contrast the following terms: statistically significant difference and clinically significant difference

Statistically significant difference o Reflects internal responsiveness and means that when scores were compared before and after treatment that there was a statistical difference o The magnitude of the difference may not be great enough for the clinician or client to find the difference meaningful Clinically significant difference o Is a change that would be regarded by clinicians and the client as meaningful and important

What is the overall purpose of regression equations?

Studies to predict outcomes use regression equations. A regression equation calculates the extent to which two or more variables predict a particular outcome. In linear regression, several predictors are entered into a regression equation to determine how well they predict an outcome of interest.

What is epidemiology?

Study of health conditions in populations. Includes descriptive research methods aimed at identifying the incidence and prevalence of specific conditions.

her notes: when would you use pearson product moment correlation statistic vs. spearman brown correlation statistic

The choice depends on how the type of variables used Pearson product moment correlation is based on 2 quantitative variables (continuous, interval) that are believed to be approximately normal in the population. Example: heart rate, blood pressure Spearman Brown correlation is based on 2 quantitative variables but one or both of them are rank ordered (or ordinal) numeric variables with the numbers not having numerical meaning. Example: stage of cancer, pain on a 1-10 scale, Likert scales

from her notes: what is the difference between validity and internal consistency?

Validity studies examine whether (or not) the totality of the measure captures the construct or what it is designed to measure. There are different types of validity. Internal consistency studies examine whether (or not), the items or subscales of the measure are correlated with the total score. Internal consistency studies help us identify if there is too much overlap in the items or if weak items are included in the measure. Cronbach's alpha and other measures of internal consistency are an 'average' of the correlations between the items and the total score (or subscores).

Describe how the measures of central tendency are affected by a skewed distribution (Fig 4-2)

When the distribution is positively skewed, the mode is a lower score than the mean, while the median falls between the mode and mean. In a negatively skewed distribution, the mode is a higher score than the mean, and once again the median falls between the mode and mean.

In a scatterplot, interpret the strength of the relationship and the direction of the relationship.

With scatterplots you can visually examine whether the relationship is weak or strong, or positive or negative. See figure 4-13- Diagonal to the right is positive, a circle is no relationship, diagonal to the left is negative. The closer the dots are to the line, the stronger the relationship.

Definition of multicollinearity

a term that refers to the circumstance in which variables are correlated with one another. in regression, it is preferable that multiocollinearity be kept to a minimum and that each predictor contribute as much unique variance as possible.

review: what does hypothesis testing provide?

an alternative approach for making inferences about population parameters from sample statistics there are also correct and incorrect ways to make conclusions about results from hypothesis testing

what is a multiple regression equation

calculates the extent to which two or more variables predict a particular outcome

Recognize differences between case-control designs, prospective cohort studies, and retrospective cohort studies.

next slides

What are the characteristics of group comparison studies and what other names are used for this type of study?

often compare a disability group to a group of individuals without a disability. these cross sectional designs that compare two or more groups at one point in time is sometimes referred to as ex post facto comparisons, they can answer important questions about the difference between groups. they are cross-sectional, non-experimental, descriptive designs that compare two or more groups at one point in time; no manipulation occurs

definition of odds ratio

probability of a statistic that determines the likelihood that, if one condition occurs, a specific outcome will also occur. odds ratio = 1 there is no difference between groups odds ratio > 1 there is a greater chance of experiencing the outcome odds ratio < 1 there is a lower chance of experiencing the outcome

Define the following terms and how they influence qualitative research: inductive reasoning

qualitative research uses inductive reasoning: data are collected and based on that data, an understanding is reached. Ex) interview a client, administer assessment, and based on the information, begin to understand

what is the repeated measures ANOVA

similar to a dependent sample or within-group t-test and is used when the means are compared over more than two time periods (i.e., the within group analysis is repeated)

Define the following terms and how they influence qualitative research: naturalistic inquiry

suggests that a phenomenon is only understood in context and that multiple perspectives can and do exist and differ among individuals gives example of being a student and if someone interviewed your classmate about their motivations for a career choice, your answers would not be the same

definition of multiple linear regression

the outcome is a continuous variable (i.e. age, severity of disorder, etc). the multiple refers to multiple predictors. the multiple linear regression reveals the total amount of variance accounted for by all of the predictors, as well as which predictors are most important

what is a mixed model ANOVA

used when between and within group analyses are conducted simultaneously; however, in the literature you will often see this referred to as a repeated measures ANOVA, with the between group analysis implied. in these analyses, two or more groups are compared over two or more time points. in a mixed model ANOVA, it is possible to examine both main effects and interaction effects.

definition of confidence intervals

useful in interpreting the meaningfulness of inferential statistics. Whenever a statistic is calculated, such as a t- or f-statistic, an odds ratio, or an effect size, the result is an imperfect estimate that contains error. The potential impact of this error can be expressed by calculating the confidence intervals of the statistic. A confidence interval is the range of values estimated for the population. A 95% confidence interval which is the most commonly reported in research studies, suggests that you can be 95% confident that the true mean of the population exists between those two values.

definition of multiple logistic regression

when outcomes and predictors are categorical, a Multiple logistic regression is used, and results are reported in terms of an odds ratio. Results given as an odds ratio Greater than 1- outcome is likely Less than 1- outcome not likely Likeliness of "if condition, then outcome"

What is a basic description of survey research and the advantages and disadvantages of this design?

- a common approach to gathering descriptive information about health conditions. a questionnaire is administered via mail, electronic media, telephone, or face-to-face contact. Used to collect large amounts of data, incidence and prevalence, and description of a phenomenon advantage: the ease of which large amounts of data can be collected disadvantage: low response rate; self-report issues (answer favorably)

from voicethread: what are statistical hypotheses?

- a hypothesis is a statement about a population parameter that can be tested -null hypothesis- the hypothesis that is tested using a test statistic based on a sample -research or alternative hypothesis: the hypothesis that you hope or expect to find true -results from the sample statistic suggest there are real differences from the statement made in the null hypothesis -conclusions may be that the research hypothesis is supported

When do you use an ANOVA test

when more than two means are compared when more than two comparisons are made, follow-up tests are necessary to determine where the differences lie.