Chapter Six Descriptive Research Designs: Survey Methods and Errors Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-1 Learning Objectives Explain the advantages and disadvantages of using quantitative, descriptive survey research designs to collect primary data Discuss the many types of survey methods available to researchers. Identify and discuss the factors that drive the choice of survey methods Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-2 Learning Objectives Explain how the electronic revolution is affecting the administration of survey research designs Identify and describe the strengths and weaknesses of each type of survey method Identify and explain the types of errors that occur in survey research Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-3 Introduction PHASE II: Design the research Marketing Research Step 3: Select the research design If the researcher requires primary data from members of the defined target population, and management’s initial decision problems are linked to specified questions concerning: Who, what, when, where and how about known members (or elements) of the target population Identifying meaningful relationships, establishing the existence of true differences and/or verifying the validity of relationships between the marketing phenomena Then the researcher should consider using a descriptive research design to collect the needed primary data Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-4 Phase II: Select the Research Design—Descriptive Research PHASE II: Design the research Marketing Research Step 3: Select the research design The need for descriptive research The desire for substantial amounts of information from enough members of the target population so that inductive logic and probabilistic inferences can be drawn Descriptive research designs include a range of survey methods: These include person-administered surveys, telephone-administered surveys, self-administered surveys and online-administered surveys Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-5 Survey Research Methods These methods are the mainstay of marketing research: Generally associated with descriptive and causal research Distinguished by the need to collect data from large groups of people Administered to selected individuals with responses recorded in a structured and precise manner Provide specific facts and estimates so that accurate predictions can be made Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-6 Advantages and Disadvantages of Quantitative Survey Research Designs Advantages of Survey Methods Disadvantages of Survey Methods Ability to accommodate large sample sizes; increases generalisability of results Ability to distinguish small differences Ease of administering and recording questions and answers Capabilities of using advanced statistical analysis Abilities of tapping into factors and relationships not directly measurable Difficulty of developing accurate survey instruments (questionnaire designs) Limits to the in-depth detail of data structures Lack of control over timeliness and, potentially, low response rates Difficulties in determining whether respondents are responding truthfully Misinterpretations of data results and inappropriate use of data analysis procedures Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-7 Types of Survey Methods— Person-administered In-home interview Executive interview Mall-intercept interview Purchase-intercept interview An interview takes place in the respondent's home or, in special situations, within the respondent’s work environment (in-office) An interview takes place with the business executive in person The interviewer stops shopping patrons and asks for feedback during their visit to a shopping mall The interviewer stops the respondent and asks for feedback at the point of purchase Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-8 Types of Survey Methods— Telephone-administered Telephone interview Question and answer exchanges that are conducted via telephone technology Computer-assisted telephone interview (CATI) An integrated telephone and computer system in which the interviewer reads the questions from a computer screen and enters respondents’ answers directly into the computer program Completely automated telephone surveys (CATS) A telephone interviewing system in which a computer dials a phone number and a recording introduces and administers the survey, leaving the subject to interact with the computer directly Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-9 Types of Survey Methods— Self-administered Direct mail survey A self-administered questionnaire that is delivered to selected respondents and returned to the researcher by mail Mail panel survey A questionnaire sent to a group of individuals who have agreed in advance to participate Drop-off surveys A self-administered questionnaire that a representative of the researcher handdelivers to selected respondents; the completed surveys are returned by mail or picked up by the representative Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-10 Types of Survey Methods— Online Fax survey A self-administered questionnaire that is sent to the selected subject by fax Email survey A self-administered data collection technique in which the survey is delivered to and returned from the respondent by email Internet survey A self-administered questionnaire that is placed on a World Wide web site for prospective subjects to read and complete See Exhibit 6.3, page 172 Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-11 Factors for Selecting the Appropriate Survey Method Situational Characteristics Budget of available resources Completion time frame Quality requirement of the data What degree of appropriate resources can be committed to the project? Consider: •Total dollars and worker hours available •Cost of collecting the required data How much time is needed for completing the research project? Consider: •Data-gathering •Analysis •Information-generation activities How accurate and representative is the derived information to the research problem? Consider: •Completeness •Generalisability •Precision Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-12 Factors for Selecting the Appropriate Survey Method Data quality determination definitions Completeness How much information and what degree of detail are needed for the defined research problem? Generalisability At what level of confidence does the researcher want to make inferences about the defined target population form the data results? Precision What is the acceptable level of error that the data results may have in representing true population parameters? Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-13 Factors for Selecting the Appropriate Survey Method Task Characteristics Difficulty of the task How much effort is required by the respondent to answer the questions? How hard does the subject have to work to answer the questions? How much preparation is required to create a desired environment for the respondents? Stimuli needed to elicit the response How much physical stimulus does a respondent need? Do specific stimuli have to be used to elicit a response? How complex do the stimuli have to be? Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-14 Factors for Selecting the Appropriate Survey Method Task Characteristics Amount of information needed from the respondent How detailed do the respondent’s answers have to be? Will probing activities be needed? How many questions should there be? How long should the respondent expect to take? Sensitivity of the research topic To what degree are the survey’s questions socially, politically and/or personally sensitive? Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-15 Factors for Selecting the Appropriate Survey Method Respondent Characteristics Diversity What is the degree to which respondents share characteristics? Consider: •Personal or behavioural characteristics •Divergence of views or opinions Incidence rate What is the percentage of the general population that is the subject of the marketing research? Consider: •Incidence rate—the relationship of the size of defined target population to general population •Impact of incidence rate on data collection costs •Accuracy of contact details Degree of survey participation How do these factors affect participation rates? Consider: •Ability to participate •Willingness to participate •Knowledge level Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-16 Types of Errors in Survey Research Methods Random sampling errors versus Non-sampling errors Sampling error Error associated with the sampled data results due to some form of natural random chance or random fluctuations in the data estimates. The statistically measured difference between the actual sampled results and the estimated true population results. Non-sampling or systematic error All errors that enter survey research design that are not related to the sampling method or sample size. Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-17 Types of Errors in Survey Research Methods Respondent error sources Refusal: A prospective respondent is unwilling to participate Non-response error •Occurs when a sufficient number of the initial prospective respondents are not included in the final sample of a study •A portion of the population is not represented or is underrepresented Not at home: A prospective respondent is not reasonably able to be reached Wrong mailing address: A prospective respondent’s address is outdated or inactive Wrong telephone number: A prospective respondent’s telephone number is no longer in service or is incorrect Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-18 Types of Errors in Survey Research Methods Respondent error sources Hostility: A response that arises from feelings of anger or resentment due to the response task Response error Social desirability: A response based on what is perceived to be socially acceptable or respectable Occurs when a significant number of respondents either unconsciously or deliberately falsify their responses Prestige: A response intended to enhance the image of the respondent Auspices error: A response dictated by the image or opinion of the sponsor rather than the actual question Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-19 Types of Errors in Survey Research Methods Respondent error sources Yea- and nay-saying: A response influenced by the tendency towards positive or negative answers Mental set error: A response based on perceptions influenced by previous responses over later ones Response error Occurs when a significant number of respondents either unconsciously or deliberately falsify their responses Extremity error: A response influenced by clarity or extreme scale points and ambiguity of midrange options Acquiescence error: A response based on the respondent’s perception of what would be desirable to the sponsor Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-20 Types of Errors in Survey Research Methods Measurement and design error sources Construct development error Incomplete constructs: Only partial data requirements are met. This creates inappropriate guidelines for scale measurement and questionnaire design activities. •Occurs when the researcher does not accurately or completely identify the important sub-dimensions of the constructs to be measured. •The data is misidentified because the composite does not represent the intended object to be measured. Low reliability or validity: Construct validity is not maintained, which increases the likelihood of collecting irrelevant or low-quality data. Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-21 Types of Errors in Survey Research Methods Measurement and design error sources Scaling measurement error •Occurs when the researcher designs inaccuracies into the various scales used to collect the primary data. •Errors could come from inappropriate questions or setups, scale attributes or actual scale points used to represent respondents’ answers. Lack of precision: Decreases reliability of data quality Lack of discriminatory power: Respondents encounter difficulties in accurate expression or sensitivity between possible answers. Data cannot be used to detect small differences Ambiguity of questions or set-ups: Misleads or confuses the respondent Inappropriate scale descriptors: Survey collects wrong or incomplete data which severely reduces the researcher’s ability to create meaningful information Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-22 Types of Errors in Survey Research Methods Measurement and design error sources Data analysis error Represents a group of errors that relate to subjecting the raw data to inappropriate analysis procedures Inappropriate analysis Predictive analysis Misinterpretation error Interpretative bias Inaccurate transformations of data structures and analysis results into usable bits of information for the decision maker Selective perception Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-23 Types of Errors in Survey Research Methods Administrative error sources Data processing error Occurs when researchers are not accurate or complete in transferring the data from respondents to computer data files Interviewer error Data coding errors Data entry errors Data editing errors Unconscious misrepresentation Occurs when interviewers distort information, in a systematic way, from respondents during or after the interviewer/respondent encounter Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso Recording errors Cheating 6-24 Types of Errors in Survey Research Methods Administrative error sources Sample design error •Represents systematic inaccuracies created by using a faulty sampling design to identify and reach the selected ‘right’ respondents •This will limit the generalisability, reliability and validity of the collected data Population specification error Sample selection error Sampling frame error Copyright 2004 McGraw-Hill Pty Ltd PPTs t/a Marketing Research by Lukas, Hair, Bush and Ortinau Slides prepared by Tony Peloso 6-25