MM551 RESEARCH METHODS CLASS

MM551 RESEARCH METHODS CLASS 3 1 CHAPTER 7 MISCELLANEOUS RESEARCH DESIGNS 2 Quantitative Approaches APPROACHES TO COLLECTING DATA • Meta-Analysis Effect Size • Baseline Designs • Reversal Designs • Multiple Baseline Designs Qualitative Approaches • Coding Systems • Content/Communication Analysis • Observational Research Time-Based Studies (Quantitative or Qualitative) • Cross-Sectional Research • Longitudinal Research 3 Typically used to summarize results from multiple experimental studies where a particular treatment is compared to other conditions QUANTITATIVE APPROACHES: META-ANALYSIS 4 Calculate a statistic called an effect size for each experimental comparison • Effect size can show the following: • How the treatment compares to other treatments • Quantify the magnitude of the treatment effect • Because effect size is standardized, the researcher can compare results across studies regardless of the original unit of measure CALCULATING EFFECT SIZE • Effect size is the difference between the groups divided by the pooled standard deviation of both groups • Weak Effect Size:= 0.20 • Medium = 0.50 • Large = 0.80+ • The greater the difference between the groups, the larger the effect size will be • An effect size close to zero means that the treatment is not any more effective than the comparison condition • A positive effect size means the treatment is more effective than the comparison condition • A negative effect size means the treatment is less effective than the comparison condition 5 This Photo by Unknown Author is licensed under CC BY-SA-NC OTHER FORMS OF EFFECT SIZE • Effect size can also be expressed in percentages and correlations • Percentage • • Correlation • • What percentage of people are better off after receiving the treatment compared to those who do not Meta-analyses can be used to combine results from nonexperimental research too Influence Analysis • Which studies have the most impact on the overall effect size 6 QUANTITATIVE APPROACHES: BASELINE DESIGNS Used if there is a limited number of subjects Works best for treatments that result in a clear and immediate effect Effects of the treatment must disappear after treatment is withdrawn • Antithesis of what we want with psychotherapy treatment, right? 7 Evaluate the impact of administering and withdrawing the treatment to see if a treatment had an effect REVERSAL DESIGN: ABA • A: Establish a baseline before administering treatment • B: Administer treatment to determine if there is a positive effect • A: Withdraw the treatment to see if problem returns Withdrawal is necessary to assure that change is due to treatment and not another factor 8 Reintroduce the treatment after the ABA sequence REVERSAL DESIGN: ABAB A: Establish a baseline before administering treatment B: Administer treatment to determine if there is a positive effect A: Withdraw the treatment to see if problem returns B: Reintroduce the treatment If reintroducing the treatment creates another positive effect, there is more confidence that the change was due to the treatment 9 Effects of treatment for psychotherapy may not disappear with treatment MULTIPLE BASELINE DESIGNS Baseline is measured for each individual in the study Treatment is staggered across individuals so nobody receives the treatment at the same time If each person improves after receiving the treatment, change is more likely due to the treatment than to outside factors that would impact subjects simultaneously See the next slide as an example 10 11 A researcher studying depression establishes the client’s baseline BDI score, then administers a new medication, then withdraws the medication, and then reinstitutes the medication. PRACTICE EXAMPLES A researcher studying depression measures several clients’ BDI scores. The researcher then introduces a new depression medication to the clients at staggered times. A researcher studying depression establishes the client’s baseline BDI score, then encourages the client to exercise for an hour per day for one week, and then instructs the client to stop exercising. A researcher is interested in determining the overall effects of exercise on depression across a number of studies, some of which use the BDI and some of which use a different measure. 12 A researcher studying depression establishes the client’s baseline BDI score, then administers a new medication, then withdraws the medication, and then reinstitutes the medication. • Reversal Design: ABAB ANSWERS TO PRACTICE EXAMPLES A researcher studying depression measures several clients’ BDI scores. The researcher then introduces a new depression medication to the clients at staggered times. • Multiple Baseline Design A researcher studying depression establishes the client’s baseline BDI score, then encourages the client to exercise for an hour per day for one week, and then instructs the client to stop exercising. • Reversal Design: ABA A researcher is interested in determining the overall effects of exercise on depression across a number of studies, some of which use the BDI and some of which use a different measure. • Meta-Analysis Effect Size 13 QUALITATIVE APPROACHES: CODING SYSTEMS Predetermined categories developed by the researcher for recording data • Frequency of appearance • Behaviors • Theories • Themes • Traits, etc. • Lack of presence Need to verify that coding system has good interrater reliability • Every researcher who participates must use the same criteria in the coding 14 CODING IN COMMUNICATION ANALYSIS Note: Communication Analysis also called Content Analysis Studies the content of recorded media or communication Good for analyzing changes over longer periods of time (culture, treatment approaches, etc.) TV, movies, speeches, publications (often journal articles) 15 Define overall purpose of study and specific research questions COMMUNICATION ANALYSIS Develop coding systems to answer each question Steps of Communication Analysis Use the coding system to analyze the media Limitation of communication analysis: May not perfectly capture what is happening currently in a field Tabulate the results 16 Different from naturalistic observation in ethnography Careful measurement and recording of specific, narrow range of behaviors in a specified setting OBSERVATIONAL RESEARCH Usually recorded by videotape and coded later so that raters can start/stop to code Uses coding system to measure behaviors rather than a narrative to describe them as in ethnography Raters need to be carefully trained Can also be used to measure key variables in quantitative forms of research, including experiments 17 Categories must be exhaustive •All possible behaviors that might be observed in setting CODING SYSTEMS IN OBSERVATIONAL RESEARCH Categories must be mutually exclusive •Ensures that raters are not confused as to which category applies Categories must be clearly defined Need to evaluate interrater reliability overall and perhaps for each category Need to verify validity of coding system •Correlate with other measures to establish criterion or construct validity See example on next slide 18 19 PRACTICE EXAMPLES Researcher creates categories to code different themes in written materials related to first generation immigrant family stress upon arrival in this country Researcher creates categories to code behaviors of the participants in a therapy session where a therapist uses EMDR with children with PTSD 20 ANSWERS TO PRACTICE EXAMPLES Researcher creates categories to code different themes in written materials related to first generation immigrant family stress upon arrival in this country • Coding in Communication Analysis Researcher creates categories to code behaviors of the participants in a therapy session where a therapist uses EMDR with children with PTSD • Coding in Observational Research 21 TIME-BASED STUDIES 22 Cross-Sectional Research Longitudinal Research Individuals of different ages are studied at the same time Collect data from the same individuals over time • Example: 3-year-olds compared to 6-year-olds • Example: At ages 3 and 6 ADVANTAGES AND DISADVANTAGES • Time to complete • Cross-Sectional: Faster • Longitudinal: Slower • Longer study takes, more problems may arise • Drop out of study • Lose contact with participants • Lose funding • Research questions may become obsolete 23 PRACTICE EXAMPLES Researcher studying the effect of depression on autism studies 10 children of a variety of ages at one time Researcher studying the effect of depression on autism studies 1 child as the child transitions through childhood 24 ANSWERS TO PRACTICE EXAMPLES Researcher studying the effect of depression on autism studies 10 children of a variety of ages at one time • Cross-Sectional Study Researcher studying the effect of depression on autism studies 1 child as the child transitions through childhood • Longitudinal Study 25 CATEGORIES OF PSYCHOTHERAPY RESEARCH Outcome Research Process Research 26 Is treatment effective and how does its effectiveness compare to alternate treatments? Types of effectiveness that can be measured OUTCOME RESEARCH • Amount of change (how much behavioral change occurs) • Economics (how much does the treatment cost) • Time (how much time does the treatment require) • Number of people (how many people are treated at once – individual, group, family) • Who administers treatment (paraprofessionals versus licensed) • Dropout rate (does one treatment lead to a higher dropout rate) • Long-Term Impact (does one treatment lead to better gains over time) 27 Experimental (Typical) OUTCOME RESEARCH TECHNIQUES Reversal Baseline Multiple-Baseline Stronger studies will look at outcomes from several different perspectives and research designs 28 PROCESS RESEARCH What processes occur in therapy? Evaluate critical events in therapy and measure whether event leads to positive outcomes • Outcome: Has there been change? • Process: How did the change occur? 29 PROCESS RESEARCH TECHNIQUES Observational Research Task Analysis • Study critical events and try to uncover patterns that lead to positive or negative outcomes 30 Self-Report (surveys) Micro-Theories MICRO-THEORIES • Process research can lead to the development of Micro-Theories • Micro-Theories can guide therapists in moment-to-moment decisions within the therapy session • Hope is to develop step-by-step directions 31 Study wants to determine if DBT is effective in treating children with PTSD PRACTICE EXAMPLES: OUTCOME OR PROCESS? Study wants to understand how a therapist’s nonverbal language affects the change process Researcher wants to develop step by step directions for therapist to follow to effect client change Researcher wants to compare the time required to achieve the same results for psychoanalysis versus CBT Study wants to determine if DBT is effective in treating children with PTSD • Outcome research ANSWERS TO PRACTICE EXAMPLES Study wants to understand how a therapist’s nonverbal language affects the change process • Process research Researcher wants to develop step by step directions for therapist to follow to effect client change • Process research Researcher wants to compare the time required to achieve the same results for psychoanalysis versus CBT • Outcome research 33 BREAKOUT SESSION Find a research article (within the last 5 years) on Autism Spectrum Disorder or Solution-Oriented Brief Therapy for couples. Complete a Critical Analysis Worksheet. CHAPTER 8 ETHICS AND VALUES IN RESEARCH 35 1932 Tuskagee Study (1932): 400 African American men with syphilis observed for 40 years HISTORY LESSONS •Never offered treatment •Ended in Class Action Lawsuit with large settlement Milgram Study (1963): Attempt to study how far subjects will go in following the directions of an authority figure to administer shocks to another person •Subjects experienced significant emotional distress while giving the shocks 1963 36 • Harm may result from: POTENTIAL FOR HARM • Generating strong emotions: Anger or fear • Significant stress • Damaging person’s self-esteem • Encouraging person to perform illegal/immoral acts • Emotional distress 37 STEPS FOR PREVENTING HARM Try Try to anticipate what type of harm may occur Estimate Estimate the likelihood that the harm will occur Evaluate Evaluate what the harm will be if it should occur Debrief Debrief the participants at the end to ensure that the participation did not cause emotional harm Exclude Exclude individuals who are most at risk for experiencing harm 38 5 RESEARCH ETHICAL PRINCIPLES 1. Demonstrate that benefits of the study outweigh the costs • Someone other than the researcher should review the study • Determine who will receive the benefits and who will incur the costs • What if study will add to body of knowledge but not directly benefit the participants? • Provide potential benefits/costs to individuals before they agree to participate in study 39 5 RESEARCH ETHICAL PRINCIPLES 2. Participation must be voluntary • Dual relationship with the participants (also serving as therapist or professor) • Large incentive can make it difficult to decline participation, especially if from vulnerable population 40 5 RESEARCH ETHICAL PRINCIPLES 3. Privacy of participants must be protected • Possible Safeguards • Confidentiality: Only researcher knows who the participants are • Anonymity: Researcher does not even know who the participants are • Reporting of aggregate data: Report only summary statistics (%, averages, medians) • No reporting of individual responses • Use pseudonyms • May not be sufficient if setting and participants are described in sufficient detail so that identities of participants can be identified • Inform participants of steps that will be taken to protect privacy • Warn participants that it may not be possible to guarantee confidentiality • Participants can then elect not to participate or carefully consider their responses 41 5 RESEARCH ETHICAL PRINCIPLES 4. Obtain informed consent • Researcher must share important information about the study with participants prior to their participating • Purpose of study • Potential benefits or risks • How privacy will be protected • For minors or impaired adults, caregivers must also consent • Using deception makes it impossible to have fully informed consent • Risk of harm must be low • Should debrief individuals afterwards to reveal the true purpose of the study and ensure no harm has occurred 42 5 RESEARCH ETHICAL PRINCIPLES 5. Ethics committees/Institutional Review Boards should approve proposed research beforehand • Most universities and organizations have a committee in place where approval is required prior to beginning the study 43 • How dishonesty occurs DISHONESTY IN RESEARCH • Alter data or create fictional data • Publish only data that supports hypotheses and withhold findings that contradict hypotheses • Conflicts of interest (industry sponsorship) 44 VALUES IN RESEARCH • Values in Research • Two questions 1. Can research be value free? 2. Should research be value free? 45 1. CAN RESEARCH BE VALUE FREE? • Some researchers believe there is an objective reality that can be discovered, independent of the observer • The values of the researcher can be separated from what is studied • Other researchers believe it is not possible to separate the observer from what is studied • Our values shape what we see and know 46 Values in Quantitative Research VALUES IN RESEARCH 47 • Quantitative researchers seldom include a description of themselves in their studies • Implies that it is not necessary to know the researcher’s background because the background does not influence the study Values in Qualitative Research • Qualitative researchers generally believe that reality is socially constructed • Shaped by who we are and social context • We are not separate from what we study • Qualitative researchers describe their backgrounds so reader can understand how the background may influence the study 2. SHOULD RESEARCH BE VALUE FREE? • Some researchers believe that researcher’s values should be kept out of the research process as much as possible to retain objectivity • Goal of research: Understand how things are • Other researchers believe that values should inform research process • Studying things the way they are perpetuates the status quo, which can be oppressive • Research should be used to promote social change • Example: Feminist research 48 CHAPTER 9 DESCRIPTIVE STATISTICS 49 • Summarize data • Convert large amounts of data into a manageable form • Types of Descriptive Statistics DESCRIPTIVE STATISTICS • Frequency distribution • Central tendency • • Mean • Median • Mode Variation • Range • Standard Deviation • Variance 50 • Shows how often different scores or values occur for a variable • FREQUENCY DISTRIBUTIONS • Frequency Table • How often different scores occur for a variable • See next slide Bar Graph/Histogram • Each bar represents the number/percentage of people in each category • Can combine categories • See two slides down 51 52 53 FOUR PROPERTIES IMPORTANT FOR FREQUENCY DISTRIBUTIONS Measures of central tendency Variability in scores Shape of the distribution Whether there are outliers 54 • Captures the value around which most of the scores are centered MEASURES OF CENTRAL TENDENCY • Mean • Median • Mode • Mean: Sum of all scores divided by the sum of the number of scores – also called the average • Usually used with ratio or interval data • Sometimes with ordinal data • Never with nominal data 55 MEASURES OF CENTRAL TENDENCY • Median: The middle value if the scores are arranged from lowest to highest • Half of the scores are above the median, and half are below • For even number of data points, median is the mean of the two middle scores • Used with ratio, interval, or ordinal data • • Never nominal data Better measure of central tendency than mean if there are extreme scores 56 MEASURES OF CENTRAL TENDENCY • Mode: The most frequently occurring score • If two or more scores tie for being the most frequently occurring score, then each is reported as the mode • Only measure of central tendency appropriate for nominal data • Also can be used with ratio, interval or ordinal data • See the next slide for examples 57 58 A researcher looks for the middle value after administering the BDI to 20 clients PRACTICE EXAMPLES The researcher looks for the BDI score that occurs most frequently among the 20 clients The researcher looks for the average score on the BDI among the 20 clients 59 A researcher looks for the middle value after administering the BDI to 20 clients ANSWERS TO PRACTICE EXAMPLES • Median The researcher looks for the BDI score that occurs most frequently among the 20 clients • Mode The researcher looks for the average score on the BDI among the 20 clients • Mean 60 3 Measures of Variability in Scores MEASURES OF VARIABILITY Range Variance Standard Deviation Range: The difference between the highest and lowest values May give a distorted picture if there are outliers 61 MEASURES OF VARIABILITY • Variance: Sum of squares divided by the number of population scores • Conceptually similar to finding the average distance of scores from the mean • When using values fr…

Do you find yourself approaching a tight assignment deadline? We have a simple solution for you! Just complete our order form, providing your specific instructions. Rest assured that our team consists of professional writers who excel in their respective fields of study. They utilize extensive databases, top-notch online libraries, and up-to-date periodicals and journals to ensure the delivery of papers of the utmost quality, tailored to your requirements. Trust us when we say that thorough research is conducted for every essay, and our expertise in various topics is unparalleled. Furthermore, we have a diverse team of writers to cover a wide range of disciplines. Be assured that all our papers are created from scratch, guaranteeing originality and uniqueness.

Write my essays. We write papers from scratch and within your selected deadline. Just give clear instructions and your work is done

Need Help Writing an Essay?

Tell us about your assignment and we will find the best writer for your paper.

Write My Essay For Me