Ch. 7 Survey Research - PSYC 2900U - Exam

10/28/24 Final Exam 2 u Date and Time: Monday, December 9, 12-2 pm...

10/28/24 Final Exam 2 u Date and Time: Monday, December 9, 12-2 pm u Location: Regent Theatre at 50 King St E (DTR100) u Exam will consist of multiple-choice and short-answer Ch. 7 – Survey Research u questions. Covers Chapters 1 through 10 (Ch. 11 is NOT on the exam). u Exam is CLOSED-BOOK. PSYC 2900U – RESEARCH METHODS u Worth 35% of your final grade. KARLA EMENO u Bring pencils, an eraser, and your student ID. u Request a deferred exam (for a valid reason) here. 1 1 2 3 4 Survey Research Lecture Overview u Survey Research – A quantitative and qualitative method with two important characteristics: u Overview of 1. Variables are measured using self-reports survey research 2. Considerable attention is paid to the issue of u Constructing sampling surveys u Most survey research is non-experimental, but surveys u Conducting can also be used within experimental research (i.e., surveys where you manipulate an IV to assess its effect on a DV). 3 4 Some Lifetime Prevalence Results from the 5 6 National Comorbidity Survey Cognitive Model of Survey Responding Lifetime prevalence* Disorder Total Female Male Generalized anxiety disorder 5.7 7.1 4.2 Obsessive-compulsive disorder 2.3 3.1 1.6 Major depressive disorder 16.9 20.2 13.2 Bipolar disorder 4.4 4.5 4.3 Alcohol abuse 13.2 7.5 19.6 Drug abuse 8.0 4.8 11.6 *The lifetime prevalence of a disorder is the percentage of people in the population that develop that disorder at any time in their lives. 5 6 1  10/28/24 Consider this example… 7 8 Context Effects Item: How many alcoholic drinks do you consume in a typical day? Response Options: A lot more than average – Somewhat more than average – Average – Somewhat fewer than average – A lot fewer than average u Context Effects – Influences that are not related to the content of the item but to the context in which the item Cognitive Processes Exam ple appears. Question W hat “alcoholic drinks” are included? Is a “typical day” a weekday, weekend, or Interpretation both? u Item-Order Effect – When the order in which the items are Might think of recent occasions involving alcohol. Might carefully try and count presented affects people’s responses. Inform ation Retrieval num ber of drinks consum ed from past week. u Response options can have an unintended effect on Mental calculation m ight m ean dividing the num ber of alcoholic drinks Judgm ent Form ation consum ed last week by seven to com e up with an average per day. people’s responses. W hat does “average” m ean in the response options? And what would count as Response Form atting “som ewhat m ore” than average? u To prevent order effects, rotate questions and response Response Editing Might believe they drink a lot m ore than average, but they change this to options when there is no natural order. “som ewhat m ore than average” to look better to the researcher. 7 8 9 10 Types of Items: Open-Ended Items Types of Items: Closed-Ended Items u Closed-Ended Items – Questionnaire items that ask a question u Open-Ended Items – Simply ask a question and allow and provide a limited set of response options for participants participants to answer in whatever way they choose. to choose from. u Ex: Is there anything else you would like to tell us about? u Ex:On a scale from 0 (no pain at all) to 10 (worst pain ever u Useful when researchers do not know how participants experienced), how much pain are you in right now? might respond or when they want to avoid influencing u Useful when researchers have a good idea of the different their response. responses that participants might make. u More qualitative in nature. u More quantitative in nature, more common. u Take more time and effort for participants, more difficult u More difficult to write, quicker for participants, easier to for researchers to analyze, participants more likely to skip. analyze. 9 10 11 BRUSCO Model for Writing Effective Items 12 Types of Items: Closed-Ended Items u For quantitative variables, a rating scale is typically provided. This is an ordered set of responses that participants must choose from. u Typically range from 3 to 11 response options (5 and 7 are most common). u Five-point scales are best for unipolar scales where only one construct is tested – Ex: Frequency (never to always) u Seven-point scales are best for bipolar scales where there is a dichotomous spectrum – Ex: Liking (dislike very much to like very much) 11 12 2  10/28/24 Writing Effective Items (cont’d) 13 14 Formatting the Survey u Response options for categorical variables should be mutually u Begin with an introduction that encourages respondents exclusive and exhaustive. to participate and establish informed consent. u Use “Other – Please specify” option when it is not possible to be exhaustive with the response options. u Present clear instructions for completing the questionnaire. u Use “Select all that apply” when appropriate u Five or seven response options are generally ideal in terms of u Start with the most important items when the participants precision for rating scales but use more when appropriate. are more interested and less fatigued. u Typically balance them around a midpoint. u Group items by topic and type (ex: same rating scale). u Can omit a middle (or neutral) response option, but it can u End with demographic items (ex: age, education, help to include it for bipolar dimensions. income). 13 14 Non-Probability Sampling 15 16 Probability Sampling u Non-Probability Sampling – Occurs when the researcher cannot specify the probability that each member of the population will be selected for the sample. u Occurs when the researcher can specify the u Convenience Sampling – The sample consists of individuals who happen to probability that each member of the population will be be easily available and willing to participate (ex: intro psych students). selected for the sample. u Snowball Sampling – Existing research participants help recruit additional u Requires a very clear specification of the population of participants for the study interest, as well as a sampling frame. u Quota Sampling – Subgroups in the sample are recruited to be u Sampling Frame – A list of all the members of the proportional to those subgroups in the population. population from which to select the respondents. u Self-Selection Sampling – Individuals choose to take part in the research on their own accord, without being approached by the researcher directly. 15 16 Probability Sampling (cont’d) 17 Probability Sampling (cont’d) 18 u Stratified Random Sampling – The population is divided into u Simple Random Sampling – Each individual in the population different subgroups or “strata” (usually based on has an equal probability of being selected for the sample (ex: demographic characteristics) and then a random sample is drawing names from a hat). taken from each “stratum.” u Most likely to involve computerized sorting or selection of u Proportionate Stratified Random Sampling – Used to select a respondents (ex: computer randomly generates phone sample in which the proportion of respondents in each of numbers from within a given geographic area). various subgroups matches the proportion in the population. u Disproportionate Stratified Random Sampling – Used to u Cluster Sampling – Larger clusters of individuals are randomly sample extra respondents from particularly small subgroups— sampled and then individuals within each cluster are randomly allowing valid conclusions to be drawn about those sampled. Does not require a sampling frame. subgroups. 17 18 3  10/28/24 19 Conducting the Survey 20 Sampling Bias 1. In-Person Interviews – Highest response rate, closest personal u Sampling Bias – Occurs when a sample is selected in such a contact with respondents, most costly option way that it is not representative of the entire population, which produces inaccurate results. 2. Telephone Surveys – Lower responses rates, still provide some personal contact with respondents, can be costly but less u Non-Response Bias – Occurs when there is a systemic than in-person interviews, not as comprehensive sampling difference between survey non-responders from survey anymore due to reduced landlines responders. 3. Mail Surveys – Less costly still, even lower response rates, u Want to maximize the response rate to address this issue. more susceptible to non-response bias u Ex: send pre-notification message, send follow-up 4. Internet Surveys – Becoming more common, not necessarily reminders, keep questionnaire short and simple, offer higher response rates than mail surveys, can be hard to an incentive achieve a sampling frame, samples likely not random 19 20 21 Upcoming Dates 22 u Tuesday, Nov. 5 th – Ch. 8 Quasi-Experimental Designs u Tuesday, Nov. 12 th – Ch. 9 Factorial Designs u Tuesday, Nov. 19 th – Ch. 10 Single-Subject Research u Friday, Nov. 22 nd – Assignment 3 due in Canvas u Tuesday, Nov. 26 th NO IN-PERSON LECTURE (slides only will be uploaded to Canvas on Ch. 11 Presenting Your Research) NO IN-PERSON TUTORIAL (virtual exam Q&A session instead) u Monday, Dec. 9 th, 12-2 pm (DTR100) – Final Exam 21 22 23 Questions about Chapter 7? 23 4