Chapter_13_Descriptive_Research

Chapter_13_Descriptive_Research - PSYC 310 Chapter 13...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: PSYC 310 Chapter 13 Research Strategies: Descriptive Research Strategies QUALITATIVE Exploratory Small N Theory building No hypothesis Narrative report QUANTITATIVE Confirmatory: you've explored the issued, now you go test your findings Large N Theory testing Hypothesis Statistical analyses Quantitative Strategies Five major strategies, grouped under three categories, according to research goal. Category I: Category II: 1. descriptive 2. correlational Category III: 3. experimental 4. quasi-experimental 5. non-experimental Quantitative Strategies Research strategies differ in level of constraint, i.e., how much control is exercised over the data collection. The more rules and the more specific & precise the method, the higher the level of constraint. High constraint - higher precision & validity for conclusions, BUT, not flexible. Low constraint - more flexible/exploratory, but be careful about conclusions (precision? validity?). Descriptive Research Measure variables as they exist naturally. They tend to be low constraint, low internal validity and high external V Emphasis is on describing the variables, not the relationship between them. Observational Survey Case study Observational Research Three main methods for collecting data: 1 Behavioral observation 2 Content analysis 3 Archival research 1. Behavioural Observation Direct observation and systematic recording of behaviors, as they occur in a natural setting. Two major concerns: 1. Presence of researcher / habituation. People will behave differently if there is an observer. You go and observe for a few days till they get used to you before collecting data 2. Subjectivity / multiple raters / trained observers / clearly defined behaviors. You have to remember that you are also bringing in your own biases about what you are observing. Something that you might find aggressive might not be aggressive to another. What you do is get multiple observers Quantifying Observations How do you convert what you see into something measurable? What are you looking to measure? What matters more, how often a tantrum happens on how long it lasts? Frequency: Duration: how often a behavior occurs how long a behavior lasts Interval: how often a behavior occurs within a specified time frame Sampling Observations Can't always observe everything all of the time. It's more practical to take "samples" of events. Time sampling: Observe everything within a specified time frame, pause to record, observe again. e.g., 10min observe, 10min record, 10min observe, 10min record... Sampling Observations Event sampling: Look for specific events within specific intervals. Ex: in the first 20 min of game you are looking for one behavior, in 2nd, other behavior, etc... e.g., interval 1, event A interval 2, event B Sampling Observations Indiv. sampling: Observe specific individuals within specific intervals. e.g., interval 1, person A interval 2, person B. How does mom vs dad interact with children Indirect Observations Content analysis: Examination of behavior/events in literature, movies, television programs, or similar media. e.g., violence in cartoons. In one hour, how much violence occurs? Indirect Observations Archival Research: Examination of historical records to measure behaviors or events that occurred in the past. e.g., number of cases in which charges were dropped. Observation Studies Three types of observational research studies: I. Naturalistic / unobtrusive observation II. Participant observation III. Contrived / structured observation I. Naturalistic Observation Observe behavior in a natural setting as unobtrusively as possible "fly on the wall". + behavior observed in real world + authentic / high external validity + observe behaviors that cannot be manipulated - time-consuming, observer influence - subjective interpretation - tells us what but not why II. Participant Observation Interact with participants or become one of them go undercover. + access to otherwise inaccessible information + participation offers unique insight + see what and also why + authentic / high external validity - time-consuming / dangerous - observer influence - loss of objectivity III. Contrived Observation Observe participants in a setting designed to illicit the behavior of interest structured observation. + don't have to wait for behavior to occur + can be done in lab or in the field - what but not why - less authentic? Survey Research Surveys can be divided into two broad categories: the questionnaire and the interview. Questionnaires are usually paper-and-pencil instruments that the respondent completes. Interviews are completed by the interviewer based on what the respondent says. Survey Research Questionnaires: sent by mail sent via email administered in person by telephone Interviews: in person (easier) by telephone Survey Research Questionnaires quick & easy large N (10-20% return) what but not why Interviews time-consuming small N what and also why Survey Research There are three major areas of consideration involved in developing a survey: 1. the question content, scope and purpose: stay on focus with participant 2. the response format: open or close ended questions? 3. the wording of the question (next slide) Question Content Demographic information: This is information that best describes the members of your sample, such as: Age Gender Marital status Ethnicity Income level Educational level Program of study Profession Living condition Languages spoken Question Content For each question you ask, you'll need to consider several points. First, you'll need to be clear about what sort of information is needed to test your hypotheses. avoid asking unnecessary questions make sure you get the information you need Question Content 1. Is the Question Necessary/Useful? Do you need to ask this question? If so, what level of detail is required? age: specific or range? income: exact amount or range? drinking: ounces or category? Question Content 2. Are Several Questions Needed? Is your main question enough or do you need further clarification? Do you smoke? Have you ever smoked? How many cigarettes per day did you smoke? Question Content 3. Is your Question Double-barrelled? Be sure that you are asking about one thing only when you construct a question. What do you think of Concordia shuttle bus service and its drivers? Question Content 4. Do respondents have the needed information? Let's say you want to ask the question: Should Isaiah Washington have been fired from Grey's Anatomy? Respondents won't be able to answer this question if they have no idea who or what you are talking about. Ask a filter question first (e.g., Have you ever watched...) before asking respondents their opinion on a matter. Question Content 5. Does the Question Need to be More Specific? Sometimes questions are too general and the information obtained is difficult to interpret. If you want to find out respondent's opinions about a specific book. You could ask: How well did you like the book? Would you recommend the book to others? Would you look for other books by that author? Question Content 6. Is the Question Biased or Loaded? One danger in question-writing is that your own biases and blind-spots may affect the wording. What do you see as the benefits of increasing school fees? What do you see as the disadvantages of increasing school fees? Question Content 7. Will Respondents Answer Truthfully? Consider whether respondents will have difficulty giving a truthful answer. Some people are sensitive about answering questions about their exact age or income. Provide response brackets to choose from: between 30 and 40 years old between $50,000 and $100,000 Response Format Survey questions can be divided into two broad types: close-ended and open-ended. Close-ended / restricted: You provide response alternatives for respondent to choose from. Open-ended: Respondent offers own answer without choosing from a list. Close-ended Questions Dichotomous: Two possible responses: Yes/No True/False Agree/Disagree Close-ended Questions Rating scale: Respondents are provided with a list of items that the must rate in order or preference or importance. Please enter your preference for the following candidates where '1' = your first choice, '2' = your second choice, and so on. _____ Stephane Dion _____ Steven Harper _____ Jack Layton _____ Elizabeth May Close-ended Questions Likert Scale: Multiple response options, polar opposites. Respondent chooses only one. Close-ended Questions Semantic Differential Scale: Multiple response options, polar opposites. Respondent chooses only one. Close-ended Questions Checklist: Multiple response options - respondent chooses as many as applicable. Please check if you have any of the following items on the computer you use most: __modem __printer __CD-ROM drive __joystick __scanner Close-ended Questions Whenever you use a checklist, you want to be sure that you ask the following questions: are all of the alternatives covered? is the list of reasonable length? is the wording impartial? is the form of the response easy, uniform? If you are unsure about having covered all of the possible responses, allow the respondent to write in any other options that may apply. Close-ended Questions Potential problems: - honesty, "social desirability": what makes you look good - response style - positive / negative bias: either I am going ot answer all questions in either a positive or negative manner - satisficing: trying to find out what the researcher wants and answers accordingly - nondifferentiation: don't care, circle responses in patters, etc... - acquiescence: answering in the way u think the researcher wants u to answer. Positively worded question, ill agree, negatively worded Q, ill disagree Open-ended Questions You ask the respondent to comment on a particular topic or issue: What is your opinion on allowing pedophiles to reintegrate into society? Please explain your answer: You can give an indication of how much detail you want by the amount of space allocated for answering and by the use of prompts. Wording of Questions You may not get the information you are looking for if a respondent misinterprets your question. Here are some points to consider in order to ensure clarity. 1. Can the question be misunderstood? What kind of headache remedy do you use? Wording of Questions 2. What assumptions does the question make? What is your favourite brand of beer? 3. Is the time frame specified? Do you think Tiger Woods will retire? Wording of Questions 4. How personal is the wording? Is the availability of off-campus parking satisfactory? Are you satisfied with the availability of off-campus parking? Do you do drugs? Do you know of students who do drugs? Wording of Questions 5. How complex is the terminology? Have you ever suffered from halitosis? Have you ever suffered from borborygmus 6. Are acronyms easily understood? What do you think of MADD's new campaign? Are you worried about writing GRE's Case Study In-depth examination of one or more individuals of interest idiographic approach. If no treatment or intervention = case history. Case Study Strengths & weaknesses: + rich in detail + can study rare and unusual people/events + can identify exception to the rule - limited external validity - subjective interpretation Case Study Meta-analysis A statistical analysis technique that integrates the results of several independent studies on a given topic - popular in last 20 years. Unlike other observational methods, this is done only after much research has been done on the topic. Answers the question, "so what do we know?" Children & TV violence Three major findings: become less sensitive to the pain and suffering of others do not fear violence nor are they bothered by violence in general more likely to become aggressive or use harmful actions towards others. Meta-analysis Method is based on "effect size" expressed as Zscores. It weighs the cumulative evidence. Quality of the results depends on the overall quality of the observations and experiments analyzed. Meta-analysis Well conducted meta-analyses allow a more objective appraisal of the evidence than traditional narrative reviews. They provide a more precise estimate of a treatment effect. They may explain heterogeneity between the results of individual studies .. ...
View Full Document

Ask a homework question - tutors are online