LECTURE 5_MRF_QUANTITATIVE RESEARCH_SURVEY_2025-2026.pdf

# Introduction to quantitative research and survey design This section introduces quantitative research methods, focusing on surveys as a data collection tool, outlining the marketing research process, and detailing the agenda for designing questionnaires [1](#page=1). ### 1.1 The marketing research process The marketing research process involves a series of steps designed to gather and analyze information to aid in marketing decision-making. These steps are [2](#page=2): 1. Establish the need for marketing research [2](#page=2). 2. Define the problem [2](#page=2). 3. Establish research objectives & questions [2](#page=2). 4. Determine research design [2](#page=2). 5. Identify information types and sources [2](#page=2). 6. Determine methods of accessing data [2](#page=2). 7. Design data collection forms [2](#page=2). 8. Determine the sample plan and size [2](#page=2). 9. Collect data [2](#page=2). 10. Analyze data [2](#page=2). 11. Communicate the insights [2](#page=2). ### 1.2 Quantitative research methods Quantitative research employs methods that involve numerical data collection and analysis. The primary methods discussed in this context are [3](#page=3): * **Surveys:** A data collection method based on questioning respondents using a predesigned questionnaire [3](#page=3) [5](#page=5). * **Experiments:** A method involving manipulation of variables to determine cause-and-effect relationships [3](#page=3). * **Observations:** While mentioned as a quantitative method, detailed discussion is deferred to a previous lecture on qualitative research [3](#page=3). ### 1.3 Questionnaire design process The design of a questionnaire is a crucial aspect of quantitative research, aiming to translate information needs into specific, answerable questions. The agenda for questionnaire design includes [7](#page=7): * Understanding measurement and scales [4](#page=4). * Question wording [4](#page=4). * Questionnaire structure [4](#page=4). #### 1.3.1 Definition and purpose of a questionnaire A questionnaire is defined as a formalized set of questions used to obtain information from respondents. Key characteristics include [7](#page=7): * It is a structured technique consisting of a series of written or verbal questions [7](#page=7). * It must convert the required information into specific questions that respondents can and will answer [7](#page=7). * It provides a standardized approach, ensuring all participants respond to identical stimuli [7](#page=7). #### 1.3.2 Questionnaire design as a systematic process Questionnaire design is a systematic process where researchers consider various factors to effectively address research objectives and minimize response error. This process involves careful planning and execution to ensure the data collected is accurate and useful [7](#page=7). #### 1.3.3 Practical exercise in survey design An example practical exercise involves designing a four-question survey to understand the TV habits of a class. This exercise would also require identifying three advantages of using surveys [5](#page=5) [6](#page=6). > **Tip:** When designing a survey, always keep the research objectives at the forefront to ensure each question contributes to answering the core research questions [7](#page=7). --- # Questionnaire design process and question wording The questionnaire design process involves systematically developing questions that accurately and effectively gather the required information from respondents. This includes carefully considering individual question content, distinguishing between attitudinal and behavioral questions, and employing best practices for question wording to ensure clarity, avoid bias, and minimize respondent burden [13](#page=13) [16](#page=16) [19](#page=19) [37](#page=37) [41](#page=41) [44](#page=44) [45](#page=45) [8](#page=8). ### 2.1 The questionnaire design process The overall process of designing a questionnaire requires a systematic approach to ensure its effectiveness in data collection [13](#page=13) [16](#page=16) [19](#page=19) [37](#page=37) [41](#page=41) [44](#page=44) [45](#page=45) [8](#page=8). ### 2.2 Individual question content Each question included in a questionnaire must serve a specific purpose and contribute to the overall research objectives [14](#page=14). * **Necessity of questions:** Every question should be assessed for its contribution to the information required for the study [14](#page=14). * Neutral questions are often best placed at the beginning of the questionnaire to establish rapport [14](#page=14). * Filler questions can be used to disguise the true purpose of the research project [14](#page=14). * Duplicated questions may be included to assess the reliability or validity of responses [14](#page=14). * **Consolidating questions:** It is important to consider whether a single question might be answered by multiple questions to gain more nuanced data [14](#page=14). * Avoid "double-barreled" questions that ask about two different things in one [14](#page=14). * Be cautious of questions that implicitly ask "why," as they can be challenging for respondents to answer accurately [14](#page=14). ### 2.3 Distinguishing between attitudinal and behavioral questions A key aspect of questionnaire design is differentiating between questions that aim to measure respondent attitudes and those that seek information about their behavior [15](#page=15). * **Behavioral questions** are designed to elicit information about past, present, or future actions of the respondent [15](#page=15). * Examples include: "Have you ever...?" [15](#page=15). * "Do you ever...?" [15](#page=15). * "When did you last...?" [15](#page=15). * "Which do you do most often...?" [15](#page=15). * "Who does it...?" [15](#page=15). * "How many...?" [15](#page=15). * "Do you have...?" [15](#page=15). * "In what way do you do it...?" [15](#page=15). * "In the future will you...?" [15](#page=15). * **Attitudinal questions** are designed to measure a respondent's opinions, feelings, beliefs, or evaluations [15](#page=15). * Examples include: "Why do you...?" [15](#page=15). * "What do you think of...?" [15](#page=15). * "Do you agree or disagree...?" [15](#page=15). * "How do you rate...?" [15](#page=15). * "Which is best (or worst) for...?" [15](#page=15). > **Tip:** Understanding this distinction is crucial for framing questions that accurately capture the intended data. Behavioral questions focus on "what people do," while attitudinal questions focus on "what people think or feel." [15](#page=15). ### 2.4 Best practices for question wording Effective question wording is paramount to ensuring the clarity, accuracy, and unbiased nature of survey responses [38](#page=38) [39](#page=39) [40](#page=40). #### 2.4.1 Use unambiguous words Vague terms can lead to varied interpretations by respondents, compromising data quality [38](#page=38). * **Problematic:** "In a typical month, how often do you shop in department stores? _____ Never _____ Occasionally _____ Sometimes _____ Often _____ Regularly" (The terms "Occasionally," "Sometimes," "Often," and "Regularly" are subjective and can be interpreted differently by each respondent) [38](#page=38). * **Improved:** "In a typical month, how often do you shop in department stores? _____ Less than once _____ 1 or 2 times _____ 3 or 4 times _____ More than 4 times" (This offers more precise and measurable categories) [38](#page=38). #### 2.4.2 Do's and Don'ts for question wording Adhering to specific guidelines can significantly improve question quality [39](#page=39) [40](#page=40). * **Do: Be focused** * **Problematic:** "How do you feel about your automobile’s navigation system?" (This is too broad) [39](#page=39). * **Improved:** "Please rate your automobile’s navigation system on each of the following features." (Followed by a list of specific features) [39](#page=39). * **Do: Be brief** * **Problematic:** "When traffic conditions are bad, do you or do you not rely on your automobile’s navigation system to find the fastest way to work?" (This is overly long and convoluted) [39](#page=39). * **Improved:** "Does your automobile navigation system help you arrive at work on time?" [39](#page=39). * **Do: Be grammatically simple** * **Problematic:** "If you needed to find your child’s best friend’s house that was over 10 miles from your house for your child to attend a birthday party, would you rely on your automobile navigation system to get you there?" (This is a complex, multi-conditional sentence) [39](#page=39). * **Improved:** "How useful is your automobile navigation system for each of the following occasions?" (Followed by a list of occasions) [39](#page=39). * **Do: Be crystal clear** * **Problematic:** "Is your automobile navigation system useful?" (Lacks specificity) [39](#page=39). * **Improved:** "To what extent would you rely on your automobile navigation system to find someone’s house?" [39](#page=39). * **Don’t: Lead** * **Problematic:** "Shouldn’t everyone have a navigation system in their automobile?" (This question suggests a preferred answer) [40](#page=40). * **Improved:** "In your opinion, how helpful is an automobile navigation system?" [40](#page=40). * **Don’t: Load** * **Problematic:** "If navigation systems were shown to help us decrease our depletion of world oil reserves, would you purchase one?" (This combines social desirability with a hypothetical scenario) [40](#page=40). * **Improved:** "How much do you think an automobile navigation system might save you on fuel?" [40](#page=40). * **Don’t: Double-barrel** * **Problematic:** "Would you consider purchasing an automobile navigation system if it saved you time, money, and worry?" (This asks about three distinct benefits simultaneously) [40](#page=40). * **Improved:** "Would you consider buying an automobile navigation system if you believed it would reduce your commuting time by 10%?" (It is recommended to ask separate questions for savings in time, money, and worry) [40](#page=40). * **Don’t: Overstate** * **Problematic:** "Do you think an automobile navigation system can help you avoid traffic jams that may last for hours?" (This uses exaggeration) [40](#page=40). * **Improved:** "To what extent does your automobile navigation system help you avoid traffic congestion?" [40](#page=40). > **Tip:** Respondents may answer questions even if they are uninformed. Well-worded questions can help mitigate this by being clear and specific, encouraging thoughtful responses rather than guesswork [17](#page=17). --- # Measurement scales and rating scale decisions This section outlines the fundamental concepts of measurement and scaling in marketing research, detailing different types of measurement scales and common rating scale configurations and decisions [19](#page=19) [23](#page=23). ### 3.1 Measurement and scales Measurement is defined as the process of determining a description or an amount of some property of an object that is of interest. Scales are instruments used to quantify these properties [23](#page=23). #### 3.1.1 Types of measurement scales There are four primary levels of measurement scales, each with increasing levels of information and analytical possibilities [24](#page=24). ##### 3.1.1.1 Nominal scales Nominal scales use labels to classify objects or individuals. They are the most basic form of measurement and do not possess any inherent order or quantitative value beyond classification [24](#page=24). * **Characteristics:** Numbers or labels are used to categorize items [26](#page=26). * **Permissible statistics:** Mode, frequency counts, percentages [26](#page=26). ##### 3.1.1.2 Ordinal scales Ordinal scales are ranking scales that allow for the ordering of objects or individuals based on a specific attribute. While they indicate relative position, they do not provide information about the magnitude of the differences between rankings [24](#page=24). * **Characteristics:** Ranks items [26](#page=26). * **Permissible statistics:** Median, mode, frequency counts, percentages, rank order correlations [26](#page=26). ##### 3.1.1.3 Interval scales Interval scales measure the magnitude of differences between objects. They have equal intervals between scale points, but the zero point is arbitrary and does not represent the complete absence of the attribute being measured [24](#page=24). * **Characteristics:** Measures the magnitude of differences; zero point is arbitrary [24](#page=24) [26](#page=26). * **Permissible statistics:** Mean, median, mode, standard deviation, variance, correlations, t-tests, ANOVA [26](#page=26). ##### 3.1.1.4 Ratio scales Ratio scales possess a true zero point, meaning zero indicates the complete absence of the attribute being measured. This allows for the calculation of ratios and proportions, providing the most comprehensive level of measurement [24](#page=24). * **Characteristics:** Has a true zero point [24](#page=24) [26](#page=26). * **Permissible statistics:** All statistics permissible for interval scales, plus geometric mean, harmonic mean, coefficients of variation [26](#page=26). > **Tip:** Understanding the level of measurement is crucial for selecting appropriate statistical analysis techniques. #### 3.1.2 Common scales used in market research Several types of scales are frequently employed in marketing research questionnaires [27](#page=27). * **Itemized rating scale:** This scale presents a series of items that respondents can rate, such as agreement, liking, likelihood, importance, or frequency of use. The Likert scale is a prominent example used for measuring agreement [27](#page=27) [33](#page=33) [34](#page=34). * **Continuous Scales:** In these scales, respondents mark a point on a continuum representing their response [27](#page=27). * **Semantic Differential Scale:** This scale uses a series of bipolar adjectives (e.g., good-bad, strong-weak) to measure respondents' perceptions of an object, brand, or product. Respondents mark their position on a scale between these adjectives [27](#page=27). > **Example:** For a semantic differential scale measuring brand image, a respondent might rate "Chipotle restaurant" on a scale between "unappealing" and "appealing," or "expensive" and "inexpensive" [33](#page=33). #### 3.1.3 Unique rating scale configurations Beyond standard scales, unique configurations can be used, such as the "Smiling Face Scale," which uses visual cues to gauge liking, especially useful for children [28](#page=28). ### 3.2 Rating scale decisions Designing effective rating scales involves several critical decisions to ensure data quality and validity [29](#page=29) [31](#page=31). #### 3.2.1 Number of scale categories The number of categories on a rating scale impacts the amount of information captured and the precision of the measurement. Typically, scales range from five to seven categories, but this can vary [32](#page=32). #### 3.2.2 Balanced vs. unbalanced scales A balanced scale has an equal number of positive and negative response options, ensuring neutrality. An unbalanced scale may have more options on one side, which can introduce bias if not carefully constructed [30](#page=30). > **Example:** > > * **Balanced Scale:** Extremely Good, Very Good, Good, Bad, Very Bad, Extremely Bad > * **Unbalanced Scale:** Extremely Good, Very Good, Good, Somewhat Good, Bad, Very Bad #### 3.2.3 Nature of scale descriptors The labels or descriptions used for scale points are important. These can be verbal descriptions, numerical points, or a combination [32](#page=32). * **Verbal descriptors:** These provide qualitative anchors for the scale points, such as "Strongly Disagree" to "Strongly Agree" [34](#page=34). * **Numerical descriptors:** Using numbers on the scale points (e.g., 1 to 5) can offer a quantifiable measure. * **Combined descriptors:** Often, scales use both verbal and numerical descriptors to provide clarity and quantitative value [33](#page=33) [34](#page=34). > **Tip:** Clearly defined and unambiguous scale descriptors are essential for consistent respondent interpretation. #### 3.2.4 Forced vs. non-forced choice A forced-choice scale requires respondents to select an answer, eliminating the option of not responding. A non-forced choice scale allows respondents to skip items or select a "neither" or "undecided" option, which can provide insights into ambivalence [31](#page=31). #### 3.2.5 Scale configurations Different visual configurations can be used. For instance, respondents might be asked to place an "X" on a blank space between anchors, circle a number, or select from explicitly defined categories like "Very Uncomfortable" to "Very Comfortable" [35](#page=35). ### 3.3 Scaling techniques Scaling techniques can be broadly categorized into comparative and non-comparative methods [36](#page=36). * **Comparative Scaling:** Involves respondents directly comparing two or more objects. Examples include paired comparison and ranking scales [36](#page=36). * **Non-Comparative Scaling:** Respondents evaluate a single object at a time. Rating scales are a primary example of non-comparative scaling [36](#page=36). --- # Questionnaire pretesting and final considerations This topic focuses on ensuring a questionnaire's effectiveness through rigorous pretesting and by incorporating essential elements like disclaimers and informed consent before its final deployment. ### 4.1 Pretesting the questionnaire A pretest is a crucial dry run of a questionnaire designed to identify and rectify any difficulties that respondents might encounter during the survey process. It is imperative that a questionnaire is not deployed in the field for the actual survey without undergoing adequate pretesting. This process should encompass testing all facets of the questionnaire, including the content of the questions, their specific wording, the sequence in which they are presented, the overall form and layout, the perceived difficulty level for respondents, and the clarity of instructions provided. Importantly, the respondents selected for the pretest should be drawn from the same population as those intended for the actual survey to ensure the feedback is relevant and representative [46](#page=46). > **Tip:** Pretesting helps to avoid costly errors in the main survey by catching issues early, saving time and resources. ### 4.2 Survey disclaimers and informed consent Before a survey is administered, it is vital to inform potential participants about the nature of the research and how their data will be handled. This is typically achieved through a disclaimer and an informed consent process [42](#page=42). #### 4.2.1 Elements of a survey disclaimer A survey disclaimer should clearly outline the context of the survey, address confidentiality and data protection measures, and formally request informed consent from participants [42](#page=42). * **Context:** This section explains the purpose of the survey, such as being part of academic studies or a specific module project [43](#page=43). * **Confidentiality and Data Protection:** It is essential to assure respondents that their responses will be treated with confidentiality. This includes stating that any personally identifiable information (like names or emails) will be removed and that responses will be used exclusively for academic purposes. The data should be aggregated and reported in a way that prevents individual identification [43](#page=43). * **Informed Consent:** By agreeing to participate, respondents are formally consenting to the terms and conditions laid out in the disclaimer. The disclaimer should also emphasize that participation is entirely voluntary and that individuals have the right to stop or exit the survey at any time [43](#page=43). > **Example:** An informed consent statement might begin with "Welcome to our online survey! Before you proceed, please take a moment to review the following information..." [43](#page=43). ### 4.3 Questionnaire design checklist A comprehensive checklist can help ensure that a questionnaire is well-designed, effective, and addresses the research objectives thoroughly. Key considerations include: * **Addressing Research Objectives:** Do the questions fully cover all components necessary to answer the research objectives, test hypotheses, and enable necessary classifications [47](#page=47)? * **Necessity of Each Question:** Is every question included essential for achieving the research goals [47](#page=47)? * **Clarity of Formulation:** Is each question clearly and unambiguously formulated to avoid misinterpretation [47](#page=47)? * **Efficiency of Questioning:** Are multiple questions needed in certain instances instead of a single, complex question [47](#page=47)? * **Minimizing Respondent Effort:** Does answering the questionnaire require the minimum possible effort from the respondent [47](#page=47)? * **Value of Open-Ended Questions:** Do open-ended questions add significant value, and are they appropriately positioned within the survey [47](#page=47)? * **Response Alternatives (Multiple Choice):** For multiple-choice questions, do the provided response alternatives encompass all possible choices, and are they mutually exclusive [47](#page=47)? * **Question Order:** Are the questions arranged in a logical and well-ordered sequence [47](#page=47)? * **Respondent Information:** Is the respondent adequately informed throughout the survey process [47](#page=47)? --- # Assignments and project deliverables This section details the assignments and project deliverables for the marketing research module, focusing on mid-term presentations, survey question proposals, and the use of Qualtrics for survey implementation. ### 5.1 Group assignment overview The group assignment involves consolidating secondary data and interview insights, proposing at least two key questions for the survey, and ensuring all students register on Qualtrics using their provided access codes. The initial deliverable for the first part of the assignment is due in week 7 [48](#page=48) [49](#page=49). ### 5.2 Mid-term presentation: first deliverable The first deliverable for the MR semester project is a 10-minute in-class presentation [50](#page=50). #### 5.2.1 Presentation dates and deadlines Presentations are scheduled for specific dates: 28 October for groups G2 & G3, and 30 October for group G1. Presentations must be submitted via email to nathalie.martin@hesge.ch the day before the scheduled session, with deadlines of 27 October (G2 & G3) or 29 October (G1) at 12:00 PM (lunchtime). Submission file names should follow the format: `Gx_TeamY_ShortTitle_midTerm.ppt` (e.g., `G1_Team5_Water Consumption_midterm.ppt`) [50](#page=50). #### 5.2.2 Content requirements The presentation content should include: * **Context and rationale:** Justification for the chosen topic's importance, supported by secondary data [50](#page=50). * **Research objectives & hypotheses:** Clearly stated objectives and hypotheses, substantiated by secondary data [50](#page=50). * **Qualitative research:** Presentation of the interview guide and thematic analysis conducted [50](#page=50). * **Quantitative research:** Proposal of 2-3 key questions for the upcoming survey [50](#page=50). * **References:** Adherence to the HEG practical guide for citing and referencing sources [50](#page=50). #### 5.2.3 Submission details In addition to the presentation file, students must submit their interview transcripts in a single file, including transcripts from all team members and their names [50](#page=50). #### 5.2.4 Presentation first slide The first slide of the presentation must clearly display: * Group number * Team number * Topic * Title of the presentation * Team members' full names (first name and SURNAME) [50](#page=50). #### 5.2.5 Evaluation The mid-term presentation is **formative**, meaning it does not carry a mark and is solely for training purposes. Constructive feedback will be provided by peers and the professor to aid student progress. Active contribution from all team members in both preparation and delivery is mandatory; any absences require justification. A feedback form is available for the mid-term presentation [50](#page=50) [51](#page=51). ### 5.3 Qualtrics account setup All students are required to register on Qualtrics using their individual access codes [49](#page=49). #### 5.3.1 Access and registration process Students should obtain their individual access code via Cyberlean. The registration process involves [52](#page=52): 1. Navigating to `heg.qualtrics.com` [52](#page=52). 2. Clicking on "Don't have an account?" on the landing page [52](#page=52). 3. Using their `@etu.hesge.ch` email address for registration [52](#page=52). 4. Entering their access code, which follows the format `Firstname123` (e.g., `Lindsay123`) [52](#page=52). #### 5.3.2 Account details Students must use their surname and first name for their account, explicitly **NO pseudonyms** are allowed. They will also need to create a password for their account [52](#page=52). > **Tip:** Ensure you have your access code ready before starting the Qualtrics registration process to avoid delays. > **Example:** If your name is John Smith and your access code is 'john123', you will use 'john.smith@etu.hesge.ch' as your email and 'john123' as your access code. You should register your account using "John Smith" and create a password. --- ## Common mistakes to avoid - Review all topics thoroughly before exams - Pay attention to formulas and key definitions - Practice with examples provided in each section - Don't memorize without understanding the underlying concepts

| Term | Definition | |------|------------| | Quantitative research | A research methodology that involves the collection and analysis of numerical data to identify patterns, make predictions, and generalize results. It often utilizes statistical techniques. | | Survey | A research method that involves collecting data from a predefined group of respondents by asking them a set of structured questions, typically using a questionnaire. | | Questionnaire | A formalized set of questions designed to elicit information from respondents. It must translate research needs into questions that participants can and will answer accurately. | | Research objectives | Specific goals that a research project aims to achieve, outlining the information the researcher intends to gather to solve a marketing problem. | | Research questions | Questions that guide the research process, stemming from the research problem and objectives, and that the survey aims to answer. | | Measurement | The process of determining a description or amount of some property of an object that is of interest to the researcher. | | Nominal scales | A type of measurement scale that uses labels or names to classify objects into distinct categories. These categories have no inherent order. | | Ordinal scales | A type of measurement scale that ranks objects based on a specific characteristic. The order of the ranks is meaningful, but the differences between ranks are not necessarily equal. | | Interval scales | A type of measurement scale where the intervals between values are equal and meaningful. These scales have an arbitrary zero point. | | Ratio scales | A type of measurement scale that has a true zero point, meaning the absence of the attribute being measured. Ratios between values are meaningful. | | Likert Scale | A popular itemized rating scale used to measure attitudes, opinions, or behaviors by asking respondents to indicate their level of agreement or disagreement with a series of statements. | | Semantic Differential Scale | A measurement technique that uses a series of bipolar adjectives to measure the respondent's feelings or perceptions about an object or brand. | | Continuous Scales | A type of rating scale that allows respondents to place a mark anywhere along a continuous line, providing a more nuanced response than discrete categories. | | Balanced Scale | A rating scale that has an equal number of favorable and unfavorable response options, typically with a neutral point in the middle. | | Unbalanced Scale | A rating scale that has more response options on one side of the scale than the other, often used when the expected distribution of responses is skewed. | | Pretesting | A preliminary test of a questionnaire conducted on a small group of respondents from the target population to identify and correct any problems before the main survey. | | Informed Consent | A process where a participant voluntarily agrees to participate in a study after being fully informed about the study's purpose, procedures, risks, and benefits. | | Double-barreled question | A question that asks about two or more issues in a single question, making it difficult for respondents to provide a clear and accurate answer. | | Filler questions | Questions included in a survey that are not directly related to the research objectives but are used to disguise the true purpose of the study or to break up monotonous sections. |