Popular in Introduction to Research Methods in Psychology
verified elite notetaker
Popular in Psychlogy
This page Class Notes was uploaded by Melanie Maino on Wednesday February 24, 2016. The Class Notes belongs to PSYC314 at Towson University taught by Brianna Stinebaugh in Spring 2016. Since its upload, it has received 11 views. For similar materials see Introduction to Research Methods in Psychology in Psychlogy at Towson University.
Reviews for PSYC314
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 02/24/16
22216 Chapter 4 Alternatives to Experimentations Surveys and Interviews Surveys and interview are used in combination with other research designs List of pros for surveys Direct access to our participants minds Useful way of obtaining information by directly asking them We can obtain a large amount of data at any given time Cheap Gather information that isn t directly observable like a behavior would be I Information on opinions I Information on thoughts I Information on people s beliefs I Information on feelings I Information on attitudes 0 We can collect data about sensitive topics that are hard to talk about face to face Anonymous way to collect data No cause and effects Low manipulation of antecedents We vary on the imposition of units Two most common survey techniques in psychology I Written questionnaires can be mailed out handed to directly or sometimes posted online I Interviews can be face to face over the phone with one specific person at a time or with a focus group 0 Steps in constructing a survey I 1 map out our research obj ectivesin the order we want to conduct it 00000 00000 1 to 2 questions research objective I 2 We want to address our imposition of units Two ways to address this 0 1 with close ended questions where you have a limited number of answers 0 2 with open ended questions which are less structured and people are free to answer the questions in any way they want a couple of sentenced Can be intertwined o A good survey will use both types of imposition of units open amp closed I 3 we must have a way to quantify our open ended answers Content analysis reviews all open ended answers evaluates them I 4 We want to keep all questions very simple and keep our participants involved we have to make sure we are using appropriate language tailor it towards a specific population 0 Define and explain You do not want to use double negativesquot No ambiguous questions No complex questions understandable One idea being expressed per question 0 Avoiding compound questions I 5 All response choices are exhaustive you need to include on your close ended questions all possible answers use other optionsvery hard to quantify so you want to be limited on these types of other questions 67 Measuring Responses 0 O O 0 Different levels of measurements Level of measurement is the type of scale that is used to measure a response 4 Scales nominal ordinal interval and ratio Nominal lowest scale I Categorizing our answers and using names as a way to divvy up our answers to questions I We aren t quantifying everything I 2 or more answers categories I used with qualitative data I no magnitude involved or strength Ordinal I Rank ordering our responses I Magnitude strength is measured in ranks I Ex unhappy very unhappy okay happy very happy Interval second to highest level I Equal intervals between each response I NO true zero point Ratio strongest in the scale I Equal intervals between each response I There is a true zero point I Ex height time weight Always want to pick the strongest scale possible How to set up a survey 0 O 1 interest right away I first question should be Relevant to your topic Easy to answer Needs to be interesting Easy to understand Should always be a close ended question 2 Keep ETHICS in mind I IRB I Nothing too personal Ways to collect survey data 00000 O Selfadministered question Mail out survey Computer or internet most common Telephone most common Interviews face to face Focus group interviews Two ways to evaluate surveys or interviews 0 Reliability consistency of answers you are obtaining and the repeatability o Validity are you asking what you intended to ask 224 Sampling 0 We are deciding who we want and where we are going to get our participants 0 Should include all members of the population you are interested in 0 Population includes people animals or objects that have at least one common characteristic or quality 0 Our sample should be as representative as possible I We should be able to make inferences about the larger population I Should re ect the larger population 0 How generalizable our results are coincides with how representative our population is 0 Two sampling approaches 8 ways to collect our sample I Probability sampling selecting participants so that the odds of them being apart of our sample are known ahead of time most preferred way of sampling most representative high external validity 4 types of Probability Sampling 0 1 Simple Random Sampling we are taking a portion of the population selecting them to be apart of our sample in an unbiased way This selection process is done through random selection when any member of the population has an equal chance of being selected to be apart of the sample I Ex 200 species you know ahead of time that each animal has 1200 chance of being selected 0 2Systematic Random Sampling variation of simple random sampling Used with SMALL populations As a researcher you list every member of that population Then you are going to select every nth determined by the size of the population and how big you want your sample to be member of that list to be apart of the sample I Ex population of 20 people but want 5 people in your sampleto determine what n is you divide 2054so 4 being your n o 3 Stratified Random Sampling this type of population is used when there are different subgroups Still using random selection to select participants to be in the sample while keeping portions of subgroups the same as they exist in the population EVERYONE is represented fairly 0 Ex 75 female 25 male psych majors when you select your sample you must keep the portions the same So that your results will still be in line with those percentages 4 Cluster Sample used with very LARGE populations Randomly selecting already existing groups within that population to represent that specific population as a whole Used when it is hard or impossible to sample individuals Ex Elementary Schools in MD it would be very hard to take that whole population Maybe want to use cluster sample and take one or two schools and have those represent all the schools Ex Socioeconomic Class take a zip code and have the specific people in that zip code represent that large group Nonprobability sampling NO random selection So this is not as representative of the larger population Low external validity Survey research is always using nonprobability sampling because it is easier 0 4 types of Nonprobability Sampling 1 Quota Sampling select sample through predetermined quotas that intend to represent the population Participants are not being randomly selected More freedom than stratified random sampling Ex 20 people in a class10 of those people were male and 10 were femaleI wanted a sample of 10 peopleWant to keep proportions similar want 5 males and 5 females But the catch is that you can select the sample however you want first 5 people in the door 2 Convenience Sampling using groups that are readily available to be part of the sample Ex comparing different research methods classesif students take the class and pass it and the success of their graduation You will use a readily available group 3 Purposive Sampling Nonrandom samplethat is selected because people re ect the purpose of the study Have in mind the purpose of your study and base that purpose on your sample Ex Depression diagnosis vs Anxiety diagnosis and affect of their memorykeeping purpose of study in mind and going to take a sample based on that purpose 10 peopletake first 5 people w anxiety and first 5 w depression I 4 Snowball Sampling used with rare of specific cases Finding participants that fit the requirements of the study and asking them for more participants to be apart of the sample Ex Need population of 18 year olds who skateboardyou would ask them to lead you to more people in that specific populationsnowball effect Methods 0 O 0 First thing mentioned is the participant You need to include how did you get the participants What sampling techniques did you use What incentives We need as much information on your sample as possible In order to replicate a study we need to make sure we know all the information of that sample as possible Knowing all of this information allows you to figure out the generalizability Chapter 5 Alternatives to Experimentation Correlational Designs and Quasi Experimental Designs Used to establish relationships of already existing behavior Used to predict behavior Correlational Designs are used to show relationships between specific antecedents and behaviorsour antecedent conditions are already existingsimply observing them and certain behaviors Advanced correlational techniques and designs to predict causality Quasi Experimental Designs model a true experiment style but are lacking one or more essential elements of a true experiment Usually what we are lacking is manipulation or random assignment When we are selecting our participants for our different treatment conditions we are basing that off already existing qualities or characteristics 3 examples as to when we would use a Quasi Experimental Design 0 We can use with different types of participants mental illness 0 Naturally occurring situations 0 Looking at uncommon or unusual events and usually comparing them to more common events hurricane Katrina Both Quasi Experimental and Correlational Designs are high in external validity Both are also high in manipulation of units restricted and limited from the responses they want to obtain from their participants Correlational Designs tend to be low in manipulation Quasi Experimental Designs tend to varyall depends on those variables that are missing