AdvertPublic Relat Rsrch Meth
AdvertPublic Relat Rsrch Meth PBRL 340
Popular in Course
verified elite notetaker
verified elite notetaker
CHMY 141N - 00
verified elite notetaker
Mrs. Elwin Parisian
verified elite notetaker
Hannah Willms II
verified elite notetaker
verified elite notetaker
Popular in Public Relations
This 7 page Class Notes was uploaded by Freeda Crona on Monday October 26, 2015. The Class Notes belongs to PBRL 340 at University of Tennessee - Knoxville taught by M. Hoy in Fall. Since its upload, it has received 14 views. For similar materials see /class/229834/pbrl-340-university-of-tennessee-knoxville in Public Relations at University of Tennessee - Knoxville.
Reviews for AdvertPublic Relat Rsrch Meth
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 10/26/15
DEVELOPING YOUR SURVEY CODING SCHEME Introductory Comments Because you will be using a statistical software package to analyze your survey results you need to develop a quotbilingual dictionary so that you can communicate to the computer in its language ie numbers the particular survey question and your respondent39s answer for that question This quotbilingual dictionaryquot for your survey is your Coding Scheme Procedure 1 Number the returned surveys with an ID numberput that number at the top of each page in case one of the sheets gets torn off This is especially important for the last page of the survey Why 2 Decide your variable names a They must be 8 characters or less beginning with an alphabet letter b Don t forget to have your survey number as a variable too ID b It39s useful to choose a variable name that somehow describes the questionmeasure 3 Decide your coding scheme the process of assigning numbers to the possible alternatives for each question for input into the computer 2 points variable name and numbers to response options It will save you a step if you put your numbers directly on your survey for the respondent to mark EXAMPLES Are you SEX 1 Male 0 Female EXAM II STUDY GUIDE 20 of your course grade Thursday November 18 2010 NOTE Makeup exams are completely at my discretion I will consider giving you a makeup exam under the following conditions 1 You have a documented emergency 1 will not accept documentation past Friday November Simply not feeling well or not sleeping well the night before the exam or not feeling at your best to study or take the test is not a Valid reason for requesting a makeup exam Q What ifl oversleep I m called into work the alarm doesn39t go off my electricity went off my neighbors were so loud thatl couldn39t study I forgot my notes when I went home for the weekend I had a fight with my significant other I got caught in traffic etc etc etc As a group textcall each other the morning of the exam to make sure no one oversleeps Watch each other s backs A The exam is administered during our designated final exam time 7 using the classroom clock lfyou miss this time for any other reason than a documented emergency you forfeit your opportunity to take the exam lfyou arrive late your exam will be collected with the rest of the class YOU MUST TAKE THE EXAM WITH YOUR CLASS 810 AM STUDENTS MAY NOT TAKE THE EXAM DURING THE 940 EXAM PERIOD 2 You or someone on your behalfifyou physically cannot must email or cill my office 97475098 and leave a message by your class time Thursday November 18 to explain why you will miss the exam 3 You have until Friday November 19 by noon to get your documentation to me You may either leave it in my mailbox in the School office 476 Communication amp Information or fax at 97472826 Having someone else drop off your documentation is appropriate ifyou are physically unable to come If you fail to meet these criteria you will not be allowed to makeup the exam Do not ask to be the exception Test Format Multiple Choice Matching Short Answer questions like the SPSS activity but based on the Backyard Burger survey Survey Research You should know the three main goals of survey research Get people to cooperate 2 Get people to give truthful answers 3 Gather meaningful information You should know the sources of bias that may be associated with survey research and some strategies you can take to reduce potential bias Random Sampling Bias can t interview your target population 2 Systematic Error or Sample Bias imperfect research 3 You should know how to encourage people to participate and ways to deal with nonresponse Guarantee its not a sales contact or request for donations Identify research organization reduces auspices bias Provide an incentive Gain their attention mail survey Live stamp and other techniques to avoid junk mail response Follow up remindersmail or phone Foot in the door techniques small request TARGET Shorter surveys What are the various ways we can assess nonresponse I Subjective Estimates weighting certain respondent s answers based on census data I Im putation Estimates assuming non respondents are like respondents only ok for a high response rate I Substitution Someone says no to your survey substitute them with someone who meets criteria Trend Analysis most common comparing 1st and 2 d waves Sampling the non respondents use a difftechnique than survey contact about 10 o of non respondents Anything in your required reading Zikmund amp Babin Chapter 7 is fair game Review the handout that compared the strengths and weaknesses of the various survey methods Strengths Weaknesses I Door to Door M0derately fast Lengthy questionnaire I Excellent respondent cooperation High in uence of interviewer I Low nonresponse rate L0w anonymity of respondent I Low possibility for misunderstanding Difficult to followup I Versatile questioning High cost Fast Very versatile questioning L0w possibility for misunderstanding Possible urban bias M0der to low cooperation M0d to long length Mall Intercept M0d to high supervision High in uence of interviewer L0w anonymity Dif cult followup M0d to high cost Telephone Interview Very fast Avg possible misunderstanding High geogr exibility G00d respond cooperation Medium non response High supervision M0derate in uence on responses M0derate anonymity Easy followup Low to moderate cost Mail surveys High geogr exibility Slowest collection High anonymity Moderate cooperation Easy followup but takes time Not versatile questioning Lowest cost High nonresponse rate No Interviewer in uence High possibility for misunderst Internet Survey Instantaneous response High possibility for misunderst Worldwide exibility Difficult followup Extremely versatile questioning No supervision of interviewer Low cost Allows use of graphics and animation 0 Make sure you Ve read the following articles 0 Consumers rebel against marketers see BB Daily Agenda 0 Judging Survey Research Quality 0 Online Polling A QuickFixfor Faulty Surveys ltlt ltltREADll Sam pling Terms SEE LAST PAGE FOR DEFINITION Squot Sampling Frame Sampling Judgment Sampling Census Population Systematic Sampling Sampling Unit Strati ed Sampling Proportional amp Disproportional Strata Periodicity Mall Intercept Approach Simple Random Sample Cluster Sampling Convenience Sampling Quota Sampling Sampling Frame Error Snowballing o What are the steps involved in determining your sampling plan 1 De ne the population Specify the Sampling Frame Specify the Sampling UnitElement ex person household business Select the Sampling Method Determine the Final Sample Size Specify the Sampling Plan 7 Solicit the Sample 0 What is the difference between probability and nonprobability sampling Which of the above methods fall under each of these broad classi cations Probability Sampling every person has a known chance of being selected Ex Simple Random Systematic Random Stratified Random Cluster Non Probability Sampling likelihood of being chosen is unknown Ex Convenience Quota Snowballing Judgment o What are the approximate sample size ranges for projections to local regional and national populations 0 Local 250 300 0 Regional 500 700 0 National 800 1000 0 If I describe a sampling scenario you should be able identify the sampling method 99539 Anything in your required reading Zikmund amp Babin Chapter 12 is fair game Research Objectives 0 Know how the Hierarchy of Effects can be used to set research objectives I AwarenessFamiliarity I Knowledge I Beliefs I AttitudesOpinions I Preferences I Behavioral Intentions I Behavior 0 Be able to identify which level in the Hierarchy of Effects each of the class survey s objectives represents Questionnaire Design 0 Know the Do s and Don ts of question writing and questionnaire design as presented in class DON T ask leading questions DON T make partial mention of alternatives DON T use emotionally charged words DON T ask ambiguous questions DON T ask double barreled questions DON T make assumptions DO ask only relevant questions DO keep questions as short as possible DO ask questions in a logical manner DO use simple conversational language DO ask sensitive or embarrassing questions at the end DO use lter questions where appropriate DO make your survey visually appealing and easy to follow 0 Know the following terms Doublebarreled Question Pivot Questionused to determine which version ofa second question will be asked logic Screening Question Funnel Techniquegeneral gt speci c gt demographic Leading Question Counterbiasing Statementan introductory statement to a potentially embarrassing question that reduces a respondent s reluctance to answer by suggesting that certain behavior is not unusual Filter Questionasked so you don t ask the respondent irrelevant questions Openended questions 0 Be able to critique survey questions and response options Know how to correct problems to reduce biases Attitude Measurement Know the following types of scales Simple Attitude choose between 2 alternatives Ex do you approvedisapprove of Categoricalconsists of several response categories often providing respondents with alternatives to indicate to indicate positions on a continuum Comparative how does x com pare to x NonComparative X is better than x strongly disagree strongly agree LikertUsually a 5 pt scale may also be 7 or 9 pt treated as an interval scaled variable strongly disagree strongly agree Semantic Differentialan interval scaled variable a series of measures that allow you to profile multiple stimuli Likert typeNumericalvariation of likert type anchored by opposites very unim portant very im portant Graphicinterval scaled ex hurt scale with faces for young children Behavioral Differentialinterval scaled ex I probably will will not buy a car in the next 12 months Ranking Grading Scaleassign a letter grade to rate a variable Forced Choice no neutral option NonForced Choiceneutral option Balanced equal number ofnegative and positive choices Unbalancedunequal number of neg and pos choices Know how we can measure past and current behavior Types of Measurement Scales Know the four basic types of scales Nominal Ordinal Interval Ratio 0 Nominal only 2 choices respondents choose one o Ordinal responses arranged in order of magnitude don t know by how much 0 Interval responses arranged in order of magnitude and distance between options are in equivalent units Ratio possesses an absolute zero point zero meaning absence of Approx how many have you eaten in the past 30 days Know what types of statistics are appropriate for each scale type Nominal frequency in terms of n or raw number and percents Ordinal frequency in terms of n or raw number and percents Interval frequency means higher order statistics Ratio frequency means higher order statistics You will be identifying scale types on survey questions like we did in class and you did on your homework and your group survey Editing and Coding You should be able to develop a coding scheme for survey questions ie developing variable names Within SPSS constraints and assigning numerical values to options You did examples in class the homework and on your group survey Descriptive Statistics Know the basic types of descriptive statistics frequencies central tendencies measures of dispersion What are the various measures of central tendency and dispersion tendency mean median and mode Dispersion varianceleaves you with units squared and std dev the larger the std dev the more diversity in the sample You should be able to interpret frequencies means and standard deviation from SPSS printouts Hypothesis Testing amp SPSS Null Hypothesis Alternative Hypothesis Hypothesis Testing Signi cance Level You should know how to articulate the null and alternative hypotheses related to particular survey questions or a particular SPSS output le null is when you assume there is variance in the sample Alternative is ifthere is more than what you assume the differences aren t due to chance You should know how to input survey data into the SPSS program how to de ne variables and how to perform the statistical procedures discussed in class In other words you should have a basic knowledge of SPSS However you do not have to memorize the dropdown menus You should understand when the ttest or a chisquare is the appropriate test statistic chi square is to test the significance ofa crosstabbreaking apart the responses to a question where it only makes sense to ask for frequencies to see if there are market segment or demographic differences T test is when you want to compare the means for a variable with two groups You should be able to interpret SPSS printouts for ttests and chisquare The decision rules for determining whether or not the test statistic is signi cant will be provided but you have to know which signi cance level on the printout is appropriate You should be able to make appropriate marketing communications recommendations based on your statistical results De nitions Sampling Frame working population Sampling Judgment Samplingsample is selected based on researcher s judgment that the person or HH is representative of the population Census Populationtarget market Systematic Samplingchoose a starting point randomly x choose every nth entry in the list Sampling Unit single element that is subject to selection in the sample person household business who does the survey represent Strati ed Sampling Proportionalsegmentstratify your sampling frame based on 1 or 2 variables Attempts to select respondents so that the sample is representative of the proportion in the population Does NOT guarantee it Disproportionalsegmentstratify your sampling frame based on 1 or 2 variables Attempts to select respondents to over represent some segment Strata Periodicity ipattern exists in the format ofthe sampling frame list based on how the list is arranged that could bias the resulting sample Mall Intercept Approach Simple Random Sampleevery unit has an equal chance of being chosen
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'