Study Guides (238,517)
Canada (115,191)
Psychology (1,813)
PSYB01H3 (82)
Anna Nagy (35)

[PSYB01] Ch.5-12 Final exam notes

25 Pages
Unlock Document

University of Toronto Scarborough
Anna Nagy

Ch.5 Observational Methods •Observations can be classified as quantitative or qualitative: -Qualitative research: Focuses on people behaving in natural setting -Collect in-depth information on few individuals/ within limited setting -Describe world in their own words, conclusion based on investigator -e.g. focus on themes emerge from discussions and manner (non-numerical terms) -Quantitative research: Focus on specific behaviours that can be quantified -Include larger samples, conclusion based upon statistical analysis of data -e.g. % of teenagers who work and way % varies by age ➣Naturalistic Observation (Qualitative) •Naturalistic observation: Also known as field work / field observation -Make observation in particular natural setting over period of time -e.g. Sylvia Scribner’s “practical thinking” •Use naturalistic observation when wants to describe how people live/experience setting •Require accurate description and objective interpretation, with no prior hypothesis ➣Description and Interpretation of Data •Naturalistic observation demands researchers immerse themselves in situation •Goal is to provide complete, accurate picture, not to test hypothesis formed for study •Keep detailed field notes on regular basis, usually with audio- and videotape recordings •To describe setting, events and persons observed, then analyzed •Good naturalistic observation report: -Support analysis by using multiple confirmations (similar events occur/reported )  Issues in Naturalistic Observation ➣Participation and Concealment: •Researcher: May whether be participant or non-participant in social setting May whether to conceal his/her purposes from other people in setting •Nonparticipant observer: Outsider, does not become active part of setting •Participant observer: Active, insider role -Can observe setting from inside, able to experience events the same way as natural participants -May lose objectivity, which necessary to conduct scientific observation, causing biases •Concealed observation: Preferable, since presence of observer may alter behaviour of subjects -Less reactive than non-concealed observation, people are not aware they’re being observed •Decision of conceal depends on ethical concerns, and nature of particular group •Informed consent may not be necessary, may given verbally ➣ Defining the Scope of Observation •May want to studying everything about a setting, but need to limit scope of observation ➣ Limits of Naturalistic Observation •Naturalistic observation cannot used to study all issues or phenomena •Most useful when study complex social settings to understand the setting/to develop theories •Less useful for study well-defined hypotheses under precisely specified condition -Field research data connection not always be scheduled at convenient time/place -Field research is very time-consuming •Negative case analysis: Observation which does not fit explanatory structure devised by research -When result of observation do not match with hypothesis -When some of the observation are not consistent, need to revises hypothesis again -Need to examines all data to make sure they are consistent with new hypothesis  Systematic Observation (Quantitative) •Systematic Observation: Refer to careful observation of one/more specific behaviours in particular setting -Interested in only few very specific behaviors, less global than naturalistic observation -Observation are quantifiable, often has developed prior hypotheses about behaviors •Bakeman and Brownlee-Videotaped in room in “free play” situation -Unoccupied: Child not going anything, simply watching -Solitary play: Play alone with toys, not interested in others -Together: Child with other child, but not occupied with any particular activity -Parallel play: Play beside other child, but not playing with others -Group play: Play with other child, including sharing toys -Rarely being “unoccupied” to engage in parallel play ➣Coding systems: •Must decide which behavior is area of interest in systematic observation •Coding system: Set of rule used to categorize observation -To measure behaviors, should be simple as possible •Family Interaction Coding System (FICS): -29 categories of interaction, focus on how children’s aversive behaviors are learned in family •System for Multiple Level Observation of Groups (SYMLOG): -Coding interaction of groups on 3 dimensions 1)Unfriendly-unfriendly 2) Emotionally expressive controlled 3)Submissive-dominant •Have advantages to use previously developed coding system, with materials available Methodological Issues ➣ Reactivity: •Reactivity: Problem of measurement in which the measure changes the behavior being observed -Presence of observer will affect people’s behaviors -Conceal by one-way mirrors / hidden microphones or cameras -Can be reduce by giving enough time for people used to presence of observers ➣ Reliability: •Reliability: Measurement reflects true score rather than measurement error -Measures are stable, consistent and precise, indicated by high agreement among raters ➣ Sampling •Samples of behavior taken over long period have more accurate and useful data than short one  Case Studies •Case study: Provide descriptions of individual, can also be a setting -Naturalistic observation sometimes called a case study, but not necessary involve naturalistic -Can use library research and telephone interviews with no direct observation •Psychobiography: Type of case study, applies psychological theory to explain life of an individual •Study may present individual’s history, symptoms, characteristic behavior etc. •”Man S with ability to recall information”: -Could remember long lists and passage with ease -Have difficulty concentrating because mental image would spontaneously appear •”Genie”: Girl who kept isolated in the room, never spoken -Lack of language skills, never developed full language ability •”R.M” : Have extensive limbic system damage -Did poorly on social exchange problem, did well on precautionary problem •Valuable in informing us of conditions that rare or unusual •Lead to development of hypotheses that can be tested using other methods Archival research •Archival research: Using previously complied information to answer research questions -Do not collect original data, but analyze existing data (e.g. statistic records) -3 types of archival research data: 1) Statistical records 2) Survey archives 3) Written records ➣ Statistical Records: •Statistical records collected by many public and private organizations •e.g. Anderson: shown relationship between temperature and violent crime ➣ Survey Archives: •Consists of data from surveys that stored on computers and available to researchers •Database become available via internet, enable researcher to analyze data online •Able to have survey archives of randomly selected national samples ➣ Written and Mass Communication Records: •Written records: Such as diaries and letters preserved by historical societies/ ethnographies •Mass communication records: include books, magazine articles, movies etc. •Archival data can used in cross-cultural research to examine aspects of social structures •Human Relations Area Files (HRAF): -Consists of anthropologists’ descriptions of many cultures  Content Analysis of Documents • Content analysis: Systematic analysis of existing documents -Requires researchers to devise coding systems, which rater use to quantify the information -Must define categories to code the information •Problem with archival data: • Desired records may be difficult to obtain •Cannot completely sure of the accuracy of information collected by others Ch.7 Asking People About Themselves: Survey Research  Why conduct surveys •Survey: Methodology of asking people to tell us about themselves -Can study relationship among variables, and how attitudes/behaviors change over time -as a complement to experimental research •California and Wiscosin: Increase negative effect, which associate with linger work hours in high school •Have assumption of people are able to provide truthful and accurate answer in use of questionnaires •Response set: Tendency to respond to all question from particular perspective -Rather than provide answers that directly related to questions -e.g. Social desiratiblity: or known as “faking good”, individual answers in more socially acceptable way •Jourard: Suggest people are more likely to lie when they don’t trust the researcher  Constructing Question to Ask ➣ Defining the Research Objectives •First researchers must determine research objectives explicitly •Require researchers to decide on type of question to ask •Survey question about: Attitudes, beliefs, behaviour , fact and demographics etc. ➣ Attitudes and Beliefs : Focus on ways of people evaluate and think about issues ➣ Facts and Demographics: •Factual question: Ask people to indicate things they know about themselves and situation -demographic information is necessary to describe the sample (e.g. age / gender)  Question Wording •Problem of survey can cause by difficulty with understanding the question -e.g. unfamiliar technical terms, vague or imprecise terms, ungrammatical sentence etc. ➣ Simplicity •Questions asked in survey should be relatively simple •Avoid jargons and technical terms which people won’t understand ➣ Double-Barreled Question: •Avoid double-Barreled: Asking 2 things at once in question ➣ Loaded questions •Loaded question: Written to lead people to respond in one way •Different wordings may lead to different responses -e.g. May less likely to say they have “raped” someone than they “forced sex” •Questions include emotionally charged words may influence way people respond -which may lead to biased conclusions (e.g. word of rape/waste/immoral etc.) ➣ Negative Wording •Avoid phrasing question with negatives •e.g. “Do you feel city should not approve the proposed women’s shelter?” •Agree with the questions = disagreement with proposal -Can be confuse people, result inaccurate answers ➣“Yea-Saying” and “Nay-Saying” •“Yea-Saying”: Tendency of respondent employ response set to agree with all questions •“Nay-Saying”: Tendency of respondent employ response set to disagree with all questions •Question Understanding Aid (QUAID): Computer program which analyzes question wordings  Response to Questions ➣Closed-Verus Open-Ended Questions •Closed-ended question: Limited number of response alternatives are given -More structured approach, easier to code -More likely to use when dimensions of variables are well defined •Open-ended question: Respondents are free to answer in any way -Require time to categorize and code for responses, more costly -Most useful when researcher needs to know what people are thinking/view the world •Schwarz: 2 approaches can lead to different conclusions ➣Number of Response Alternatives •With closed-ended question: Fixed number of response alternatives •e.g. simple answer of “yes or no”, or 7-point scale  Rating Scale •Provide “how much” judgment on any number of dimensions (e.g. amount of agreement) ➣Graphic Rating Scale •Graphic Rating Scale: Require continuous 100-mm line that anchored with description at each end -Ruler is then placed on line to obtain score from 0-100 ➣Semantic Differential Scale •Measure of meaning of concepts that developed by Osgood •Respondents rate any concept one bipolar adjectives using 7-point scales •Concepts are rated along 3 basic dimensions: 1) Evaluation (adjectives) 2) Activity 3) Potency (weak-strong, hard-soft) ➣Nonverbal Scale for Children • Point to the face that shows how you feel about the toy ➣Labeling Response Alternatives •Respondents decide meaning of the response alternatives •Researcher may provide labels to clearly define meaning of each alternative •Schwarz: Named the a scale as “high-frequency scale” / “low-frequency scale” -most alternatives indicate a high frequency of exercise  Finalizing the Questionnaire ➣Formatting the questionnaire •Printed questionnaire should appear attractive and professional •Neatly typed, free of spelling errors, use scale format consistently •Roberson and Sundstrom: Obtain highest return rates when important questions present first -and when demographic questions asked last •Better to group questions together when they address similar theme/topic ➣Refining Questions •Better to give questions to small group of people, have them “think aloud” while answering •”Think aloud” procedure: Need to ask individuals to tell you how they interpret each questions -Provide information that can use to improve questions  Administering Surveys •Two ways to administer surveys: 1) Use written questionnaire 2)Use interview format  Questionares: •Questions are presented in written format, respondent write their answers •Advantages: Generally less costly than interviews -Allow respondents be completely anonymous, as long as no identifying information •Disadvantages: Require respondents be able to read and understand question -May find boring to set by themselves reading and providing answer ➣Mail Surveys •Surveys can be mailed to individuals at home or business address •Potentially low response rates, no one is present to help if person becomes confused ➣Internet Surveys •Easy to obtain samples of people with particular characteristics •Problem: Have ambiguity about characteristics of individuals providing information for study -People may misrepresent their age, gender, or ethnicity  Interviews •Involves interaction between people, people often more likely to answer for real person than mailed •Advantages: Good interviewers become skilled in convincing people to participate -People more likely to leave questions unanswered in questionnaires than interview -Interviewer can clarify any problems the person might have •Problem: -Interviewer bias: -Interviewers could subtly bias respondents answers by show approval/disapproval answers -Describes all biases that arise from fact that interviewer is human interact with another -Interviewer may have expectation to “see what they looking for” in respondents’ answers ➣Face-to-Face Interview •Require interviewer and respondent meet to conduct interview •Interview tend to be quite expensive and time-consuming •More likely to be used when sample size is fairly small ➣Telephone interviews •Less expensive than face-to-face interview •Computer-assisted telephone interview system(CATI): -Lower cost for labor and analysis cost, data are entered directly into computer ➣Focus Group Interviews •Interview with group of about 6-10 individuals brought together for 2-3 hours •Usually have sort of monetary / gift incentive to participate •Questions tend to be open-ended, asked of whole group •Usually prefer conduct at least 2-3 discussion g groups  Survey Designs to Study Changes over time •Surveys mostly study people at one point in time •Panel study: Study changes over time, same people are surveyed at 2 points in time -Important when research question addresses relationship between one variable  Sampling from a population •Sampling: Participants from population of interest •Population: Composed of all individuals of interest to researcher •Statistical theory: Allow us to infer what population is like, based on data obtained from sample ➣Confidence Intervals •Accuracy is within 3 % point when using 95% level of confidence •Confidence interval: -Confidence interval gives information about likely amount of sampling error -Confidence that true population value lies within interval around obtained sample result •Obtained result may deviate from true population due to sampling error ➣Sample size •Large sample are more likely to yield data that accurately reflect true population value •Sample size needed for sample % to be accurate within +/- 3%/5% and 10% given 95% confidence  Sampling Techniques •Two basic techniques for sampling individuals from population: 1) Probability sampling: Each member of population has specifiable probability of being chosen 2) Non-probability sampling: Do not know probability of particular member of population being chosen ➣Probability Sampling • Simple Random Sampling: Every member of population has equal probability of being selected -e.g. telephone interviews •Stratified Random Sampling: Population is divided into subgroups (Strata) -Random sampling techniques then used to select sample members from each stratum -Advantage of built-in assurance that sample will accurately reflect composition of subgroups -When represent small group within population, researcher will ‘oversample” that group -to ensure a representative sample of group is surveyed •Cluster sampling: When obtaining list of all member of population is difficult -Identify “cluster” of individuals, then sample from these cluster -A “multistage” approach, requires series of samples from larger to smaller clusters -Do not have to sample from lists of individuals to obtain truly random sample of individuals -e.g. classes are clusters of students  Non-probability Sampling •Only little effort expended to ensure sample accurately represents to population •Non-probability samples are cheap and convenient •3 types of sampling: 1) Haphazard sampling 2) Purposive sampling 3) Quota sampling ➣Haphazard Sampling •Haphazard Sampling: or “convenience” sampling -could called a “take-them-where-you-find-them” method of obtaining participants -e.g. Select sample of students from school in anyway that is convenient -Likely to introduce biases into samples, not an accurate representation of population of all ➣Purposive Sampling •Purposive Sampling: -Purpose: to obtain sample of people who meet some predetermined criterion -e.g. ask customers to fill out questionnaire about one/more movies ➣Quota Sampling •Quota Sampling: Choose sample that reflect numerical composition of various subgroup in population -Similar to stratified sample procedure, but without random sampling -Make sure of particular % in subgroups, but collect using haphazard techniques  Evaluating Sample •Sample should be representative of population from where they are drawn •Can be bias from 2 sources: 1) Sampling frame used 2) Poor Response rates ➣Sampling Frame •Sampling Frame: the actual population of individuals from random sample will be drawn •Need to consider how well sampling frame matches the population of interest ➣Response Rate •Response Rate: The percentage of people in sample who actually completed the survey -Indicate how much bias there might be in final sample of respondents •Lower the response rate, greater the likelihood that biases may distort the findings -Limit the ability to generalize the findings to population of interest Ch.8 Experimental Design  Confounding and internal Validity •Researcher manipulates the independent variable, to create groups that differ in levels of variable -then compares groups in terms of their scores on dependent variable •All other variables are kept constant through: 1)Direct experimental control 2) Randomization •Cofounding variable: Variable that varies along with independent variable -Occurs when effects of independent variable and uncontrolled variable are intertwined -Cannot determine which of the variable is responsible for observed effect -Would not be a factor if the variable is held constant in both conditions •Good experimental design = eliminate possible confounding that results in alternative explanation •Internal validity: When results of experiment confidently attributed to effect of independent variable -Must conduct experiment that only independent variable can be cause of results  Basic Experiment •Simplest experimental design has 2 variables: 1) Independent variable: Have 2 levels of Experimental groups Control group 2) Dependent variable •Experimental method involves control over extraneous variables -through either keeping such variable constant (Experimental control) -or using randomization to make sure extraneous variable will affect both groups equally •Simple experimental can take one of 2 forms: 1) Posttest-only design 2) Pretest-posttest design ➣Posttest-Only Design •Posttest-Only Design: Must obtain -2 equivalent groups of participants -Independent variable -Measure the effect of independent variable on dependent variable •Procedure of posttest-only design: 1) Choose participant and assign them into 2 groups -Must achieve equivalent groups to eliminate any potential selection differences - Selection differences: Difference in type of subjects who make up each group in design -Occurs when participants elect which group they are to be assigned to -People selected to be in condition cannot differ in systematic way -Achieve by randomly assigning participants / having same participants in both condition 2) Choose 2 levels of independent variable: -e.g. experimental group that receive treatment + control group which does not -or use two different amounts of independent variable 3) Effect of independent variable is measured -Same measurement procedure is used for both groups -Statistical significance test would used to assess difference between the groups ➣Pretest-posttest Design •Pretest-posttest Design: Pretest is given before experimental manipulation is introduced -Make possible to ascertain that the groups are equivalent at beginning of experiment •Usually not necessary if participant have been randomly assigned to 2 groups •In large sample participants, random assignment will produce virtually identical group -Larger the sample, less likelihood the group will be differ in any systematic way -More likelihood difference between groups on dependent variable is due to independent variable •Minimum of 20-30 participants per condition ➣Advantages and Disadvantages of 2 Designs •Randomization with small sample sizes, groups may not be equal -Pretest enables researcher to assess whether the groups were in fact equivalent •Sometimes pretest is necessary to select participants: -Need to give pretest to find lowest or highest scorers on measure -Can measure extent change in each individuals -Necessary whenever there’s possibility that participants will drop out the experiment -Able to assess effects of mortality, to see does it affected the final results •Mortality: Dropout factor in experiments -People may drop out for reasons unrelated to experimental manipulation •Disadvantage in Prestest: -Can be time-consuming and awkward to administer -Can sensitize participants to what you’re studying, enable them to figure out the hypothesis -May then react differently to manipulation •Pretest can be disguised by administer it in completely different situation with different experimenter -To embed pretest in set of irrelevant measure, then not obvious that researcher is interested -Assess impact of pretest by conduct combination of both posttest-only/pretest-posttest -Solomon four-group design: Half receive only posttest, other half have pretest-posttest -Posttest scores will be the same in 2 control groups if no impact of pretest  Assigning participants to experimental conditions •Independent groups design: Participants are randomly assigned to various conditions -Each participates in only on group •Repeated measures design: Each participants assigned to both levels of independent variable -Each participant is measured after receiving each level of independent variable ➣Independent Groups design •Different participants are assigned to each of condition using random assignment •Random assignment: Prevent any systematic biases, group will be equivalent -Participant differences cannot be explanation for result of experiment  Repeated Measures Design •Same individuals will participate in both conditions •Participants are repeatedly measured on dependent variable after being in each condition of experiment ➣Advantages and Disadvantages of Repeated Measures Design Advantages: •Fewer research participants are needed in repeated measure design •Extremely sensitive to finding statistically significant different between groups -since have data from same people in both conditions •Individual differences can be seen and explained Disadvantages: •Different conditions must be presented in particular sequences -There’s greater recall in high-meaningful condition •Order effect: Order of presenting treatments affects the dependent variable -Performance on second task might improve merely due to practice effects -Associated with passage of time: -Practice effect: Improvement in performance as result of repeated practice with task -Fatigue effect: Deterioration in performance as research participant becomes tired/bored •Time-related order effects are possible whenever there’s sequence of tasks to perform •Effect of first treatment may carry over to influence response to second treatment •Contrast effect: Occurs when response to second treatment in experiment is altered -Because 2 conditions are contrasted -Solve by employ counterbalancing techniques -Sole by devise procedure which interval between condition is long enough Counterbalancing ➣ Complete Counterbalacing •Counterbalacing: All possible orders of presentation are included in experiment -Possible to determine the extent to which order is influencing results ➣ Latin Squares •Use latin squares to control for control effect without having all possible orders •Latin square: Limited set of orders constructed to ensure: 1) Each condition appears at each ordinal position 2) Each condition precedes and follows each condition one time  Time Interval Between Treatment •To counterbalancing the treatment order: -Researchers need to determine time interval between presentation of treatment •Rest period may counteract a fatigue effect  Choosing between Independent groups and Repeated Measures Designs •Two advantages in Repeated measures: 1)Reduction in # of participants required to complete experiment 2) Greater control over participants differences, greater ability to detect effect of independent variable •Procedure that produce relatively permanent change in individual cannot use repeated measures Matched design •Another method of assigning participants to conditions •Matched pair design: First match people on participant characteristic, not using random assignment -Ensure that the groups are equivalent -More likely to be used when only few participants are available, very costly and time cosuming -Have greater ability to detect statically significant effect of independent variable -Worth only when matching variable is strongly related to dependent measure •Matching variable: Either the dependent measure/ variable that strongly related to dependent variable •Goal to achieve same equivalency of groups that achieved with repeated measures design -Without necessity of having same participants in both conditions •1) Obtain a measure of matching variable from each individuals 2) Participants are rank ordered from highest to lowest based on scores on matching variable 3) Form matched pairs that are approximately equal on characteristic 4) Members of each pair are randomly assigned to conditions in experiment Ch.9 Conducting Experiment  Selecting Research Participants •Sample may drawn from population using probability sampling / non-probability sampling •Use probability when it’s important to accurately describe the population •Use non-probability sampling when research is testing hypotheses -Focus of study is relationship between variable/test prediction derived from theory -Can use non-probability haphazard / “convenience” sampling •Larger sample size = result will statistically significant = more accurate estimate of population  Manipulating the independent variable •Need to construct operational definition of variable to manipulate the independent variable -Need to turn conceptual variable into set of operation (e.g. instructions/events/stimuli) •”Setting the stage”: Independent and dependent variable must introduced within context of setting ➣ Setting the stage •Two things to do in setting the stage: 1)Provide the participant with informed consent information 2) Explain to participants why experiment is being conducted •No clear-cut rules, but setting must seem plausible to participants Types of Manipulations ➣ Straightforward Manipulation •Manipulate variable with relative simplicity by presenting written, verbal material to participants •Straightforward manipulation: -Manipulate variable with instructions and
More Less

Related notes for PSYB01H3

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.