Research 2

baseline phase

"A" refers to the ____________ of a single N design

manipulation phase.

"B" refers to the ____________ of a single N design

Participatory Action Research (PAR)

"Democratic" research method Growing area of research that is often qualitative, but may be quantitative Involves a team - researchers, service users (clients/patients), clinicians - all with equal power Results are more likely to be trusted Aims to improve services and empower clients May use a combination of above approaches

How do we use conduct qualitative research?

...comparing people's, groups, or population perspectives ...exploring the meanings, symbols, beliefs, or values of a person or culture ...gathering multiple perspectives of people or populations ...observing people's actions, engagement, or participation

Purposes of Qualitative Research

1. May explore an understudied phenomenon. Attempts to discover key constructs about the phenomenon 2. May attempt to find out what or how something needs to be studied. To examine previous assumptions Describe experiences Document processes

Qualitative MethodsData Gathering

1. Participation 2. Observation 3. Interviews 4. Document review 5. Focus groups 6. Other

Sampling in qualitative studies

2 main considerations: Appropriateness. Choosing people that will best inform the researcher about the phenomena Adequacy. The amount of data will fully describe phenomena of interest - "Saturation" Consider Maximum variation Homogeneity Theory-based selection

Case study or Single N Design?

A case study should be chosen when you: Want to gain a holistic sense of a case Have questions about how or why a phenomenon occurred but do not have the ability to control variables. Are using primarily qualitative measures that primarily assess past occurrences A single N study should be chosen when you: Want to examine a specific cause-effect relationship Have questions about how a manipulation impacts an individual Are using quantitative measures that can be repeated on a daily or weekly basis

Grounded theory

A method to build a theory from case study data is

Chi-square goodness of fit

A nonparametric test used with one nominal variable having two or more categories; tests whether the observed frequencies of the categories reflect the expected population frequencies

Chi-square test for independence

A nonparametric test used with two nominal variables having two or more categories; tests whether the frequency distributions of two variables are independent

Ha: There will be a difference in verbal aggression after playing 45 minutes, 30 minutes, 15 minutes of no minutes of a violent computer game.

A researcher wants to examine the effect of violent computer games on verbal aggression. She has college students play 45 minutes, 30 minutes, 15 minutes or no minutes of a violent computer games. Which of the following best represents a non-directional alternative hypothesis for this study?

predict future behavior.

A stable baseline is important in order to

Multiple-baseline across settings

A therapist develops a new treatment to help a client quit smoking. The client collects baseline data on how many cigarettes he smokes each day in his car, at work, and in his house The therapist then tells the client to implement the treatment in his car, and then a week later in his house, and then finally at work. What type of design is this?

What does a Reversal Design for Single N designs look like?

ABA, ABAB

Different Flavors of ANOVA

ANOVA examines the variance between groups and the variances within groups These variances are then compared against each other Similar to t Test...only in this case you have more than two groups One-way ANOVA Simple ANOVA Single factor (grouping variable)

0

According to the null hypothesis for a simple experiment, we expect the difference between the means of our two groups to be:

Case Study

An in-depth examination of a single individual, group, event, or organization is called a

False

An independent t-test involves at least 3 groups.

Single N Designs

Another alternative to a sample-based study is the single N design. Quantitative investigation of a cause and effect relationship within a single case. Small N designs: A series of single N designs. Used to determine if results from one single N study generalize to other subjects or participants

Issues to Consider in Matched Pairs Designs

Are there one or two participant variables that might confound your results? Can you find an appropriate matching variable? Does the matching variable sensitize the participants to the purpose of your study? Does matching drastically reduce the available participants? How much time and effort does the matching process require?

Issues to Consider in a Repeated Measures Design

Are there several participant variables that are likely to be confounded with the effect of the DV? Does your IV create a permanent change in participants? Will multiple measures decrease the internal validity of your study through sensitization, carryover, practice, fatigue/boredom or response sets? Does your study require lengthy time period for the data collection, making your results vulnerable to participant history or maturation effects? Is attrition a problem due to the time commitment required of participants?

Assumptions for Chi goodness of fit

Assumptions: One variable with two or more categories (or groups) The categories are independent (there is no matching or repeated measures) An expected frequency (E) of at least 5 in each category Every member in the analyzed dataset belongs to only one of the categories Analysis: Compare the observed frequency (O) to the expected frequency (E)

Assumptions for Chi test for independence

Assumptions: Two variables with two or more categories Independent groups (no matching or repeated measures) An expected frequency (E) of at least 5 in each cell Every member in the analyzed dataset belongs to only one of the cells Analysis: Compare the observed frequency (O) to the expected frequency (E) for each cell

order effects.

Counterbalancing eliminates

Conducting a Single N Design: what is it composed of?

Baseline (A phase): In a single N design, repeated assessment of the dependent variable in the absence of any manipulation. Manipulation (B phase): In a single N design, repeated assessment of the dependent variable during the implementation of a manipulation (e.g., treatment). Visual Inspection: A non-statistical technique in which patterns of the A and B phases are compared.

False

Both qualitative and quantitative research studies are concerned with the reliability and validity of measurements.

2 ´ 3

Carlos is examining whether parenthood (yes - no) is related to willingness to support a bond for schools (yes - maybe - no). He could present the data in a __________ contingency table

True

Chi-square goodness of fit assumes an expected frequency of at least 5 in each category

How to address order effects in Designs with More than Two Dependent Groups ?

Complete counterbalancing: counterbalance all possible sequences. Partial counterbalancing: each condition is represented in each order but not all possible sequences are represented. Latin square counterbalancing: A type of partial counterbalancing where each condition appears once in each sequence. Randomized partial counterbalancing: Randomly assigning each participant to one of the possible sequences of conditions without concern about order or sequence

How do we Analyze an Independent-Groups Design?

Confidence interval: Defines the interval that we are confident contains the population µ difference represented by our sample mean difference; typically we compute the 95% confident interval Squared point-biserial correlation (rpb2): A measure of effect size for the independent-samples t test, providing the percentage of variance in the outcome (or DV) accounted for by the predictor (or IV). Consider also the practical significance of the results.

Correlational Design

Correlational design: a type of study that tests the hypothesis that variables are related, with no manipulation of variables. Predictor: Variable that is used to predict the value of another variable, and a term used instead of IV in a correlational design. Outcome: The variable that is predicted, and a term used instead of DV in a correlational design. Example of a correlational design comparing two independent groups: -Predictor = Chocolate-eating habits (does the participant eat chocolate daily or not?) -Outcome = IQ score

contingency table

Data for a chi-square test for independence are presented in a

Qualitative Research

Definition: A research approach that explores a social or human problem. Aims to be holistic and capture complex issues. Characteristics: Results summarize analysis of words, pictures, or observations about a phenomenon Often quite detailed Conducted in the "natural setting"

df = (n1 - 1) + (n2 - 1)

Degrees of freedom in for an independent-samples t test is computed by:

True

Dependent-groups designs allow you to ensure that relevant personal characteristics are equalized across groups at the beginning of a study.

Matched pairs and repeated measures designs are analyzed in the same way, using what?

Dependent-samples t test: 2 dependent groups Also called a paired-samples t test and within- subjects t test Effect size = rpb 2 or Cohen's d

Education level is related to membership in the investment club.

Dr. Welthe examines whether the number of those with a college degree and those who have not earned a degree who are members of an investment club differs. She finds that χ2 obt = 6.54 and χ2crit = 3.84, and can conclude:

Saturation

During Jane's McDonald's study, she interviewed 5 young workers. She analyzed interviews 1-3 and created a list of codes. However, when analyzing the 4th worker's interview, she found that no new codes were identified that she could add to the list. She verified this with the 5th interview, when again no new codes were identified during analysis. Jane is experiencing

One-Way Within-Subjects ANOVA

Effect size = partial eta squared (η2partial ) "Partial" indicates that the variability unique to individual participants has been removed. If F is statistically significant, run post hoc tests. Fischer's LSD In SPSS, use Bonferroni Adjustment

Ways to Design a Simple Experiment

Ensure IV manipulation is reliable and valid If possible, choose a manipulation that has been used in past research Administer IV in the same way (consistency of treatment) Consider including a manipulation check Ensure DV measure is reliable and valid If possible, use an existing measure or scale Be sure the measure will be sensitive to the IV manipulation Use more than question or item to operationally define the DV Automate or standardize data collection Use multiple measures to assess DV

One-Way Between-Groups ANOVA tests

Eta squared (η2): The statistic used to assess the effect size in studies analyzed with an ANOVA Post Hoc Tests: Test performed after you obtain a significant overall F with 3 or more groups; the results tell you which groups differ from one another Fisher's Least Significant Difference (LSD): A commonly used post hoc test that computes the smallest amount that group means can differ in order to be significant.

Which qualitative theory fits this scenario? Research question: ...to explore and describe strategies used by care partners to support and maintain participation in community mobility, ...considering participation of care partner dyads as a social-interpersonal unit. Sample: 3 care dyads, all long-term married partners, retired Methods: repeated, semi-structured interviews with both care partners present, participant observation in community mobility activities, photographs of the community sessions, and reciprocal data analysis. Results: overall, strategies were more relational than procedural. Reported community mobility profiles & shared community mobility themes including values of the relationship, finding/using available technology, and relying on social networks.

Ethnography

Experimental Design

Experimental design: a type of study that tests the hypothesis that an IV causes a change in a DV. The IV is manipulated and participants are randomly assigned to condition. Simple Experiment: An experiment in which the IV has two conditions and the DV is measured on an interval or ratio scale. Example of a simple experiment: -IV = Chocolate eaten before a test (students are randomly assigned to receive a chocolate treat or to not receive a treat) -DV = test scores

Phenomenology

Explores the questions: "What is it like to...?" "How is the ordinary experienced?" And, "What is the meaning in it?" Focuses on an individual person's experiences, or the variety of people's experiences in regards to a specific phenomena sampling: Intentional Purposive To gather range of lived experiences Data: Collection: Some observation Mostly in-depth conversations Analysis: No specific structure Should be tailored to the data

How to Interpret ANOVA results?

F (2,27) = 8.80, p < .05 F = test statistic 2,27 = df between groups & df within groups {Ah ha...3 groups and 30 total scores examined} 8.80 = obtained value Which we compared to the critical value p < .05 = probability less than 5% that null hypothesis is true Meaning the obtained value is GREATER than the critical value

why is the ANOVA test an Omnibus Test?

F test is an "omnibus test" and only tells you that a difference exist Must conduct follow-up t tests to find out where the difference is... BUT...Type I error increases with every follow-up test / possible comparison made 1 - (1 - alpha)k Where k = number of possible comparisons

One-way Between-Samples ANOVA

F= Between group variance -------------------------------- Within group variance Between-groups variance: Variability in scores created by the different levels of the IV Within-groups variance: Variability in scores created by individual or participant differences

More Complicated ANOVA

Factorial Design More than one treatment/factor examined Multiple Independent Variables One Dependent Variable Example - 3x2 factorial design

what are advantages of a case study?

Gain in-depth information about a case or set of cases. Useful in examining rare phenomena Usually involves collecting qualitative data that provides a holistic sense of a case

Correlational and Experimental Designs:

Goal for both is to examine a relationship Correlational designs cannot examine causation Experiments examine causal relationships These designs use inferential statistics to test if a relationship found within a sample represents a relationship that exists in the population. A larger sample increases the researcher's ability to find a statistically significant result in the sample when that result exists in the population (i.e., power).

Descriptive Design:

Goal is to describe characteristics of a population. Ability to reach this goal depends on how representative the sample is of the population. You are more likely to have a representative sample if you: Use probability (random) sampling Have a large sample

Which qualitative theory fits this scenario? Research question: 'When faced with an ethical issue in practice how do occupational therapists come to an ethical decision that can be enacted within their practice context?' Sample: practicing occupational therapists in Canada, n = 18 Method: In-depth interviews in person or over phone, internet Data analysis: 1. initial coding, 2. focused coding, 3. theoretical categorization/sorting, 4. theory construction Findings: the resulting categories from coding

Grounded Theory

Qualitative Approachesaka Traditions

Grounded Theory Phenomenology Ethnography Participatory Action Research There are more, but these are the most common to occupational therapy.

Assumptions for dependent samples t test

IV (or predictor) is dichotomous (nominal scale with two groups) Groups are dependent via matching or repeated measures Because the groups are dependent, the n of the two groups is always equal DV (or outcome) is interval or ratio scale of measurement DV (or outcome) is normally distributed

Matched Pairs Design (dependent design)

Identify a matching variable that you expect is highly correlated with the outcome (or DV). The matching variable might be pretest scores on the outcome or DV, or another variable that is expected to impact results. It should not, however, be the same as the predictor (or IV). Create pairs based on scores on the matching variable. In a correlational or quasi-experiment, match pairs across the pre-defined groups. In an experiment, each member of the pair is randomly assigned to the IV condition. Measure the outcome/DV and compare scores among the pairs.

a practice effect.

If the participants' performance improves as a result of performing the same task repeatedly, this is known as

retain the null hypothesis.

If tobt = −1.98 and tcrit= 2.16, we would

Participation in qualitative study

Immersion in setting Goal: To help identify what is important to learn about To discover issues/events that might not be obvious to an outside To look past preconceived ideas & assumptions Most often used with: Ethnography Research must be reflective & self-aware

Improving Rigor

In Study design: Collect data over an extended period of time Use a variety of data collection methods Triangulation. When 2 or more people who have independently observed & recorded their own notes. In analysis - double-check data & patterns: Use multiple coders or analyst to review the same data Have results reviewed by study subjects (member-checking) Verify results using previous study results Internally reflect on the findings, and if they make sense

N by the number of categories (k)

In a chi-square goodness of fit the expected observations are computed by dividing

1

In a study examining whether a person lives alone (yes or no) is related to their compliance with a home exercise program (complied or did not comply), what would be the df

decrease; homogeneity

In her study of adjustment to college, Sasha decided to use only first year, first generation college students who lived on campus. This decision is likely to ___________ the error variance in her results because of the _________ of her sample.

True

In independent t-test designs the participants in each level of the IV variable are unrelated.

.01

In order to be 99% confident you have not committed a Type I error, at what level should you set your p value?

Advantages of multiple comparisons within a single study:

Increases efficiency Decreases chance of Type I error Decrease confounds Increases external validity by examining functional relationships

Analyze an Independent-Groups Design

Independent-samples t test: 2 independent groups Assumptions IV (or predictor) is dichotomous (nominal scale with two groups) Groups are independent DV (or outcome) is interval or ratio scale of measurement DV (or outcome) is normally distributed Variability (SD) in each sample is similar (homogeneity of variance)

How different from independent-samples t test?

Independent: Goal is to compare the overall difference between the two groups Calculate M & SD using individual scores in each group Dependent: Goal is to compare the difference between scores from matched pairs (or within participants) Calculate the M & SD of differences between scores rather than original scores.

Grounded Theory: What is it?

Inductive method - starts with the data (minimal literature review) Creates conclusions (results) from the data gained from the study to explain the phenomenon of interest Good for developing theories about processes that occur in specific settings, such as clinical Sample sizes: generally between 20-50 participants Negatives: Reduced ability to generalize results as previous studies are not taken into consideration Data Collection: Formal structure for data collection Uses: Narratives Focused interviews Informal discussion Participant observation Field notes Theory Development: Analysis: Codes and categories are derived from the data alone "Zig-Zag process" Results: A final theory is presented, based on analysis

Writing a Qualitative Article

Introduction Literature review/background May be brief Methods Include sample, data collection & processing procedures, analytic strategies, potential bias/limitations Findings (results) Broad & specific, using examples from original data Discussion Summarizing results, directing future research, describing theory generated from study

Unstructured interview

Jane asked the manager, George, at the setting she was researching "What do you think the main differences between older and younger workers are?" and they conversed about this subject for the next 10 minutes. This is an example of:

Hawthorne effect

Jane is observing workers at McDonalds to assess the work behaviors of younger and older people. Before her research study, she would come to this McDonalds and noticed that the older people were often a bit more reserved, but more engaged than the younger people. Now that she has received an approval to observe and had an HSIRB signed, she notices that everyone is more engaged with the customers and sometimes will glance her way. What may Jane be experiencing?

Observation

Jane is sitting at a table at McDonalds, watching people order and the staff prepare and deliver the food. She is taking notes, recording information, as she watches. What data collection method is she using?

we have violated our assumption of homogeneity of variance.

Levene's test tells us if

Ordinal Data with Independent Groups tests

Mann-Whitney U test: One variable with two independent groups and n < 20/group; For effect size compute Rank Sums test, ignoring the requirement that one group must have n > 20 Rank Sums test: One variable with two independent groups and n > 20 for at least one group Kruskal-Wallis H test: One variable with three or more independent groups and n > 5 for each group, allows unequal n/group; Post hoc test is rank sums test Effect size is version of eta squared (η2)

Designing a Simple Experiment

Maximize Power Strong manipulation of the IV Extreme levels of the IV Homogeneity of participants Increase N Maximize Internal Validity Eliminate confounds: conduct a pilot study reduce demand characteristics by conducting a single-or double-blind experiment.

Nominal Data with Dependent Groups test

McNemar's test: Two dependent categories with matched or repeated measures Cochran Q test: Three or more dependent groups assessed on a dichotomous variable; If significant, the post hoc test is McNemar's test using Bonferroni's correction

Other Types of Single N Designs

Multiple Baseline The manipulation is introduced at different times across two or more persons, settings, or behaviors. Provides more control and the ability to see if results generalize. Multiple Manipulation Two or more manipulations are implemented over the course of the study. Allows the researcher to compare effectiveness of different manipulations.

have ordinal or nominal data

Nonparametric statistics are used when we

Chi-square tests (χ2):

Nonparametric tests used with nominal data which compare expected vs. observed frequencies

Degrees of Freedom (df) for ANOVA

Numerator Number of groups minus one k-1 3 groups --- 3 - 1 = 2 Denominator Total number of observations minus number of groups N-1 100 participants - 30 - 3 = 97 Represented: F (2, 27)

Document Review in qualitative studies

Often completed on pre-existing information (secondary data) Ex. Diaries/journals, Clinical documents, Newspapers/magazines, Social media, Photographs Unobtrusive Considerations: Selecting appropriate documents Gaining permission Confirming meanings through additional methods

degrees of freedom

The value used in an independent t-test that approximates the sample size is known as:

alpha

Type 1 error is also called

Ethnography

Often used in Anthropology Researcher works to gain the "insider perspective" Explores cultures; tells group stories about daily life & cultural meanings, beliefs, and social patterns May also explore organizations, programs that focus on common social problems to document how they work, the behaviors associated with them Critical ethnography: includes an advocacy perspective

what is a Case Study?

One alternative to a sample-based study is the case study. Detailed investigation of a single individual, group, organization, or event. Multiple case studies Compare patterns across cases. Embedded case study: Investigation of single cases that comprise a group or organization in order to understand that group or organization as a whole.

False

One of the best reasons to use a case study is to test a causal relationship

How do we Analyze a Multiple Independent-Groups Design?

One-way between-subjects ANOVA: 3+ independent groups Assumptions IV (or predictor) has 3 or more levels (conditions) Groups are independent (participants belong to only one level/condition) DV (or outcome) is interval or ratio scale of measurement DV (or outcome) is normally distributed Homogeneity of variances (variability in each sample is similar)

Analysis of 3+ Dependent Groups

One-way within-subjects ANOVA: 3+ dependent groups Also called repeated measures ANOVA and dependent-groups ANOVA Assumptions: IV (or predictor) has 3 or more levels (conditions) Groups are dependent (matched or repeated measures) DV (or outcome) is interval or ratio scale of measurement DV (or outcome) is normally distributed Sphericity in variances of the differences between pairs of groups

Which qualitative theory fits this scenario? Research question: ...to evaluate the strengths and weaknesses of an informal education program (IEP) in a minimum security prison and to identify what elements of the program were valued by participants. Method: PAR with 1 clinician and 3 inmates met 3-4x/month for 6 months to plan. 27 former participants of the IEP program volunteered to be interviewed using a semi-structured approach. Results: organized using 2 phases of summative content analysis methodology - 1st, a count of frequencies of words & concepts in the transcripts; 2nd, an interpretation of words/concepts to better understand meaning. High frequency words: 1. Doing, 2. Information, 3. Re-entry fears (socialization), 4. Technology, 5. Self-worth.

PAR

Parametric statistics vs. nonparametric statistics

Parametric statistics: Used to analyze interval and ratio data Assumptions include a normal distribution and homogeneity of variance Typically requires at least 10 cases Nonparametric statistics: Use to analyze nominal and ordinal data Use when the assumptions of parametric statistics are violated

Repeated Measures Designs (dependent design)

Participants serve as their own control and experience all conditions. They complete multiple measures of the outcome/DV Scores on the outcome/DV are compared within each participant Very powerful design, but subject to order effects An order effect occurs when it is possible that the order of conditions impacted the results, rather than the predictor/IV To minimize order effects, counterbalance so that participants receive conditions in different orders. In an experiment, each participant is randomly assigned to the order of conditions.

Daily exercise will increase happiness

Peter is studying the effect of daily exercise on happiness. His directional alternative hypothesis might state:

Which qualitative theory fits this scenario? Research question: The study focuses on experiences of existing environmental control systems (ECS) users with cervical SCI. Sample: 5 males with cervical SCI who used smart-device ECS Method: Semi-structured interview, 20 open-ended questions Results: 2 main categories, 1. opportunities, 2. costs. Opportunities had 6 themes (independent control, choice, peace of mind, connection, effective resource use, control over phone and applications). Costs had 4 (financial, time, frustration, technical limitations).

Phenomenology

Effect size statistics/tools for Chi test for independence

Phi squared (φ2): The statistic used to assess the effect size for a 2 X 2 Chi-square Test of Independence. Contingency coefficient squared (C2): Used to determine the effect size for a contingency table larger than 2 x 2 and with an equal number of rows and columns (3 x 3, 4 x 4, etc.) Cramer's V squared (V2): The effect size statistic used when a contingency table is larger than a 2 x 2 and the number of rows and columns are different numbers (3 x 4, 4 x 2, etc.)

Advantages of single N designs

Potential to examine cause and effect Valuable supplement to randomized experiments Track progress on a more frequent basis Repeated assessment allows for flexibility Useful in examining rare phenomena

Quasi-Experimental Design

Quasi-experimental design: a type of study that tests the hypothesis that an IV impacts a DV. The IV is manipulated but participants are not randomly assigned to condition. Example of a quasi-experimental design comparing two independent groups: -IV = Chocolate eaten before a test (chocolate is given to students in one class but not given to students in another class, so the IV is manipulated but students are not randomly assigned to group) -DV = test scores

Computing the F Statistic: what do we want for the variances?

Rationale...want the within group variance to be small and the between group variance large in order to find significance.

what are limitations of a case study?

Relies on anecdotal information that is subjective and difficult to verify Limited ability to determine causality due to a lack of control Details of a case study can be so persuasive so results are weighed more heavily than they should be. Debate as to whether or not multiple case studies can be used to build, expand, and generalize theories. Proponents of grounded theory promote the development of theories based on multiple case studies. Others argue that cases cannot be generalized to theories due to their anecdotal nature.

Data Analysis in qualitative studies

Remember the overall goal: To analyze "data" (text, observations/field notes, verbal content) into the variety of meanings to answer the research question. Depends on data gathering method used Typically inductive, or starts with the data and organizes it into codes and/or categories

rule out alternative explanations for causality.

Repeated assessment is used in a single N design in order to

Disadvantages of single N designs

Repeated assessment requires a considerable amount of time and effort. The researcher must choose an assessment technique that yields reliable data The assessment and manipulation must be followed consistently, while avoiding other systematic variations. Ability to generalize to other individuals or to support or develop a theory is questionable.

Limitations of Sample-Based Designs

Results do not necessarily apply to individual cases. Sample-based studies average results across participants May exclude certain types of participants in order to increase power or for ethical reasons. Always a chance for error, so that results cannot generalize to every individual in a population.

Rigor in a Qualitative study

Rigor: increasing confidence in study results by Following rules, procedures, and/or techniques That have been developed & agreed upon by scientific community Use evidence-based procedures Be precise and detailed Can be addressed in study design or in analysis Consider saturation

Observation in qualitative studies

See and hear events, behaviors, objects in the setting Record data using field notes Researcher needs to be objective, unobtrusive May contribute rich data to study Participant observation Mix of participation & observation "Active observation"

Ordinal Data

Spearman's rho (rs): used to analyzes the relationship between two ordinal variables. Interpret like a Pearson's r: The value of the statistic varies from -1.0 to +1.0 The sign (+ or -) indicates the direction of the relationship The absolute value indicates the strength of the relationship

stable baselines versus unstable baselines for Single N designs

Stable baseline: A baseline that displays no trend (or slope) and little variability and therefore allows for prediction of future behavior. Unstable baselines make it difficult to interpret results. Baseline is in opposite direction to what is expected during the manipulation phase (least problematic) Baseline is in the same direction as what is expected during the manipulation phase Baseline is variable

Pre-Analysis Steps for qualitative studies

Step A: Data Processing Transforming data into permanent records Makes a copy of original data Transcribing: The process of listening to taped recordings and typing verbatim what is said Step B: Break down copied information into "data units" Units: Blocks of information that will be analyzed together May be several page stories, paragraph of text, or phrase/term used repeatedly Compare/contrast data over data units Look for how/why the data is important to the people interviewed/observed

graphing data and determining patterns using visual inspection.

The analysis of a single N design typically involves

homogeneity of variance.

The assumption that variances in populations are the same is called

an IV that is dichotomous, DV that is interval or ratio, independent groups, homogeneity of variance.

The assumptions for the independent-samples t test include

the mean difference.

The average difference between the scores of matched pairs or the scores for the same participants across two conditions is known as

observed frequencies with expected frequencies.

The chi-square test compares

the frequency distributions of two variables are independent

The chi-square test for independence tests whether

what is a Dependent-groups design?

The design reduces error variance and is more powerful than an independent design Matched pairs design: A design where participants in each group are matched on a characteristic relevant to the variable that is being measured Repeated measures design/within subjects design: A design where participants experience every condition in a study

False

The expected frequency for red jellybeans in a jar of colored jellybeans is the actual number of red jellybeans

False

The independent-samples t test is used to test the differences in means in a study with a sample mean and a population mean.

one-way ANOVA.

The inferential statistic used to analyze a multiple independent-groups study is

independent-samples t test.

The inferential statistic used to analyze two-group designs is called the

0; no differences between the observed and expected frequencies

The lowest possible value of χ2 is ____________, which represents __________.

average deviation within all groups of the study.

The mean square within groups represents the

independent

The null hypothesis for a chi-square test for independence predicts that the two variables are

between-groups variance and within-groups variance

The one-way ANOVA compares

Sample-Based Designs

The primary goal of these designs is not to just describe the sample or examine the relationships among variables within the sample. Rather, the primary goal is to extrapolate results to the population from which the sample was drawn.

Use of triangulation during the data collecting

The quality of a qualitative study may be improved by

the means of the two groups in our study

The sampling distribution for an independent-samples t test is a distribution of the differences between

AB

The simplest single N design possible is

dependent-groups design.

The type of study that best supports the goal of beginning with similar participant characteristics in our groups is the

Why do we Base Study on Past Research?

Use past research to help you define and refine your topic. Use past research to help determine what research questions most warrant further investigation. Use past research to help determine the best research design.

Analysis of Variance (ANOVA)

Used when more than two group means are being tested simultaneously Group means differ from one another on a particular score / variable Example: DV = GRE Scores & IV = Ethnicity Test statistic = F test R.A. Fisher, creator

Interviews in qualitative studies

Verbal exchange In person, Over the phone, Skype Non-verbal communication may be observed & documented Considerations: Developing the interview guide Presentation & listening skills Ability to use probes when needed Documentation of discourse In-depth interviews Unstructured Conversational, casual with general guide More open ended questions Probes used to elicit more information Semi-structured & structured Mixture of fixed-response and open-ended questions May have categorical answers Interviewers ask questions precisely as written Focus groups Small group (4-6 people) Conversation in supportive setting focused on a fixed topic

True

We may be able to extrapolate results from a sample to a population, but we cannot necessarily extrapolate results from a sample to an individual.

Grounded Theory

What qualitative approach would be best suited to use to answer the following research question: "How may therapists support people with with long-term, complex medical needs get necessary equipment that support participation, specifically wheelchair(s)?"

Phenomenological

What qualitative approach would be best suited to use to answer the following research question: "What participation barriers do wheelchair users experience in their community?"

Random selection

What sampling method should not be used in a qualitative research study?

Single N design

What type of study is the best choice when you want to examine a specific cause-effect relationship for a specific individual and are using quantitative measures that can be repeated on a daily or weekly basis

Maximum variation

When Jane selected the interviewees for her McDonald's study, she chose a Caucasian 19-year old female who was high school dropout and a single mom, a 16-year-old African American male who was a junior in high school, an 18-year-old Hispanic female who was just graduating from high school, and a 19-year-old Asian male who was in his first year of college. Based on the information here, what is the most likely sampling strategy that Jane used?

Type II error

When you accept a false null hypothesis you are making a

Cohen's d

Which is an appropriate measure of effect size for dependent-samples groups?

Larger sample size

Which of the following is NOT a characteristic of a dependent-groups design when compared to an independent-groups design?

The groups are independent.

Which of the following is NOT an assumption of the dependent-samples t test? Question options: -The predictor is dichotomous. -The n of the groups is equal. -The groups are independent. -The outcome is interval or ratio.

whether the client has a had a previous stroke

Which of the following variables would be appropriate for a Chi-square: -whether the client has a had a previous stroke -body mass index -how many cigarettes a person smoke per day -number of siblings in the family with previous stroke

Ordinal Data with Dependent Groups tests

Wilcoxon T test: One variable with two dependent (matched or repeated) groups; No effect size Friedman χ2: One variable with three or more dependent groups, if 3 conditions requires n > 10/group, if 4 conditions requires n > 5/group; Post hoc test is Nemenyi's procedure (similar to Tukey's HSD) Effect size is version of eta squared (η2)

True

You do not need to compute a post hoc test or effect size for a chi-square goodness of fit test

Single N design

a quantitative design used to examine a cause and effect relationship within a single case.

Adding additional phases in a single N design

helps to clarify the relationship between the manipulation and the dependent variable.

Do not use a repeated-measures design if you expect:

the condition will change participants in some permanent way. completing multiple measures will sensitize participants to the study's expectations. a carryover effect: the effect of one condition of the treatment continues (or carries over) into the next condition. a practice effect: participants' scores change due to repeating a task rather than because of the level of the IV a fatigue effect: changes in the DV occur because of participants becoming tired

x2(2) = 20.6, p < .05

x2 represents the test statistic 2 is the number of degrees of freedom 20.6 is the obtained value p < .05 is the probability