Quantitative Research Dr. Judy Li Zhu The Questionnaire ‘A structured data collection mechanism involving a range of question formats and completed orally or in print. Questionnaires may be administered by interviewers or selfcompleted by the respondent.’ (CIM, 2003) 2 Questionnaire Design: two-way communication (Wilson, 2006) 3 The Questionnaire Design Process (McGivern, 2006) Decisions about question content, wording and order are the end result of a process that considers the following: • What is the research problem? • What types of evidence are needed to address it? 4 The Questionnaire Design Process (Wilson, 2006) 5 The Questionnaire Design Process Step 1: Develop question topics • Research objectives Which topics? Are they of primary/secondary importance? Classification information? How will the results be analysed, presented and used? • Qualitative research findings To clarify what should be asked in the questionnaire To determine relevant wording and scales • Characteristics of respondents 6 The Questionnaire Design Process Step 2: Select question and response formats Types of Response Format • Closed ended • Dichotomous question • Multiple choice question • Open-ended • Unstructured question • Projective techniques: word association, sentence completion, story completion and picture completion. 7 Scaling • Nominal scales allow us to place an object in one and only one of a set of mutually exclusive classes with no implied ordering. • Ordinal-scale data involves ranking. • Interval-scale data reflects how much more one object has of an attribute than another object. It is possible to tell how far apart two or more objects are with respect to the attribute. • Ratio-scale data possess the same kind of properties as interval-scaled data but also possess an absolute or natural origin. Scale Types • Comparative scaling: the respondent is asked to compare one set of objects against another. • Non-comparative scaling (monadic scaling): respondents are required to evaluate each object independently of other objects being investigated. Comparative Scales • • • • • Paired comparisons Rank-order scales Constant sum scales Line marking/continuous rating comparative scale Q-sort scale Comparative Rating Scales In comparison to the eggs that you have previously bought from Tesco, rate the quality of the eggs from Asda using a rating scale where 1 is significantly lower quality and 5 is significantly higher quality. Much lower Lower quality About the Higher quality Significantly quality than than Tesco same quality than Tesco higher quality Tesco than Tesco 1 2 3 4 5 12 The Questionnaire Design Process Step 2: Select question and response formats Commonly used scaling approaches: • Constant sum scales • Likert scales • Semantic differential scale 13 Constant Sum Scales Below are five characteristics of food processors. Please allocate 100 points among the characteristics in a manner that represents the importance of each characteristic to you. Characteristics of food processors Number of Points Has many accessories Is made by a well-known manufacturer Is easy to clean Is stylish Is quiet to operate Total: 100 points 14 An Example of Likert Scale Items I would now like to find out your attitudes towards this model of phone. Therefore for each of the following statements, please tell me if you strongly agree, agree, neither agree nor disagree, disagree or strongly disagree. Strongly________________________Strongly Agree Disagree The phone is easy to use The phone’s design is stylish The keys are too small I would look good with a phone like this I don’t need a phone like this 15 An Example of a Semantic Differential Scale (Wilson, 2006) 16 Non-Comparative Scales • E.g. itemised rating scale. • The scale has numbers and/or a brief description associated with each category and the respondent is asked to select one of the categories, ordered in terms of scale position, that best describes the object under study. • Itemised rating scales are often employed to measure purchase intentions. Purchase intention scales represent an attempt to measure a respondent’s interest in a brand or product. Non-Comparative Rating Scale Rate the quality of the eggs that you have previously purchased from Asda using a rating scale where 1 is very poor quality and 5 is very high quality. Very poor quality 1 2 3 4 5 Very high quality 18 Single-item vs. Multiple Item Scales • A multiple item scale consists of a number of statements to which the respondent must react. For example, a respondent might be asked to indicate how favourable or unfavourable each statement is. An overall score is then determined by combining the reactions to each. Attitude Measurement • Projective Methods: to enable respondents to indirectly ‘present data’ about themselves by speaking through others. • Word association: words matching • Story completion: partial scenarios presented • Role playing: participants are asked to assume the role of another person and to behave in a manner in which they feel that person would act. • Lists: lists of various products or activities are shown to participants. They are asked to describe the type of person who would use such a list. Attitude Measurement Self-reporting methods • Semantic differential scale: a number of bipolar adjective phrases and statements that can be used to describe the objectives being evaluated from unfavourable to favourable or vice versa. • Staple scale: how well one term describes the subject. • Likert Attitude scales: statements with which the respondent has to agree or disagree are rated on the basis of the strength of their personal feelings toward it. The Questionnaire Design Process Step 2: Select question and response formats Evaluating the validity and reliability of scales: Reliability refers to the extent to which the scale produces consistent results Validity of a scale relates to whether the differences in the scores on the scale reflect true differences in what is being measured 23 The Questionnaire Design Process Step 3: Select wording Avoid… • • • • Ambiguous wording Double-barrelled questions Leading or loaded questions Implicit assumptions 24 Questionnaire Structure • The introduction – it must be persuasive, and it usually must qualify the respondent as someone who belongs in the sample. • The body or content – consists of questions that cover information needed to solve the marketing problem – facts, knowledge, opinions and attitudes, motives, and possible future behaviour. • Basic data – includes demographics about the household, aspects such as family size and nature, income, etc., demographics about the respondent as well, and it may include questions about lifestyle and psychographics. Question Contents • • • • Is the question necessary? Does the respondent understand the question? Will the question elicit the required data? Will changing the order of the words produce different answers? • Is there more than one question being asked? • Does the respondent have the necessary information to answer the question? • Is the respondent willing or able to answer the questions? Question Phrasing • • • • • • • • • • Clarity and simplicity Length of questions Ambiguity and vagueness Avoid biased words Avoid leading questions Avoid negative questions Avoid questions asking for estimates or generalisations. Avoid hypothetical questions Avoid implicit assumptions Response choices should not overlap. The Questionnaire Design Process Step 4: Determine sequence Funnel sequence of questioning (Wilson, 2006) 28 Question Sequence and Layout • Earlier questions should attempt to create interest in the respondent. Questions should follow a logical order. • Questions relating to age, salary, etc. should be left until last, except where use is being made of a quota sampling in which case it is necessary to establish this information at the outset of the interview. • Layout: easy to use and encourage the user to answer questions A questionnaire which is badly presented or printed is likely to cause non-response or measurement error. The Questionnaire Design Process Step 5: Design layout and appearance The key elements are… • • • • Spacing Quality of production Variety Coding/analysis requirements 30 The Questionnaire Design Process Step 6: Pilot test • The questionnaire should be administered in the same manner as for the full survey • Involve interviewers (if appropriate) in pilot testing • Thorough pilot testing is critical to the success of a survey 31 The Questionnaire Design Process Step 7: Undertake survey • After pilot test, managers, who are using the information, should approve the questionnaire before it goes ‘live’ • As data are collected, input responses into appropriate computer software (e.g. SPSS) 32
Purchase answer to see full attachment