π Lecture 18 β Designing a good research questionnaire...contd
π Overview: This lecture continues the discussion on designing effective research questionnaires, focusing on choosing the structure of questions (unstructured vs. structured) and the critical task of wording questions properly. It provides guidelines for avoiding common pitfalls and establishing a logical sequence for questions, which are essential for collecting valid and reliable data.
ποΈ Topics Covered
The lecture first differentiates between unstructured (open-ended) and structured (multiple-choice, dichotomous, scale) questions, detailing their advantages and disadvantages. It then provides extensive guidelines for choosing question wording, including defining the issue, using ordinary and unambiguous words, and avoiding leading questions, implicit alternatives, and assumptions. Finally, it offers suggestions for sequencing questions, covering the funnel approach, placement of sensitive topics, and the order of basic, classification, and identification information.
π Lecture Summary
Choosing Question Structure
A question may be unstructured or structured. Unstructured questions are open-ended, allowing respondents to answer in their own words. Structured questions specify a set of response alternatives and a response format, such as multiple-choice, dichotomous, or scale.
Unstructured Questions
Unstructured questions (also called free-response or free-answer questions) allow respondents to answer in their own words, e.g., "What is your hobby?". They are good as first questions on a topic because they provide rich insights and are useful in exploratory research. A principal disadvantage is the high potential for interviewer bias. Another major disadvantage is that coding of responses is costly and time-consuming.
Structured Questions
Structured questions specify the set of response alternatives and the response format.
Multiple-choice Questions
In multiple-choice questions, the researcher provides a choice of answers, and respondents select one or more alternatives. The response alternatives should include the set of all possible choices and be mutually exclusive. Order or position bias is the tendency to check an alternative merely because of its position; each alternative should appear in each extreme position once.
β Advantages: Reduces interviewer bias, quick to administer, and less costly/time-consuming for coding. β Disadvantages: Considerable effort to design, difficult to obtain information on alternatives not listed, and potential for order bias.
Dichotomous Questions
A dichotomous question has only two response alternatives (e.g., yes/no, agree/disagree), often supplemented by a neutral alternative like "no opinion" or "don't know". A key issue is whether to include the neutral alternative.
π Example: "Do you intend to buy a new house within the next three months?" with options "Yes," "No," and "Don't know."
π Guideline: If a substantial proportion of respondents can be expected to be neutral, include a neutral alternative. If the proportion is expected to be small, avoid it.
Scales
This topic was discussed in detail in previous chapters (8 and 9).
Choosing Question Wording
Deciding on question wording is the most critical and difficult task in developing a questionnaire. The following guidelines help avoid problems.
Define the Issue
A question should clearly define the issue being addressed. For example, "Which brand of soap do you use?" is incorrect because it is vague. A better wording is, "Which brand or brands of soap have you personally used during the last month? In case of more than one brand, please list all the brands that apply."
Use Ordinary Words
Ordinary words matching the vocabulary level of the respondents should be used. Technical jargon should be avoided. For example, "Do you think the distribution of soft drinks is adequate?" is incorrect. A better wording is, "Do you think soft drinks are readily available when you want to buy them?"
Use Unambiguous Words
Words like "usually," "normally," "frequently," and "often" can have different meanings. Specific options should be used. For example, instead of "Never," "Occasionally," "Sometimes," "Often," and "Regularly," use specific ranges like "Less than once," "1 or 2 times," "3 or 4 times," and "More than 4 times."
Avoid Leading or Biasing Questions
A leading question clues the respondent to what answer is desired. For example, "Do you think that patriotic Pakistanis should buy imported automobiles when that would put Pakistani labor out of work?" is biased. A more neutral wording is, "Do you think that Pakistanis should buy imported automobiles?"
Avoid Implicit Alternatives
An implicit alternative is not explicitly expressed in the options. For example, "Do you like to fly when traveling short distances?" has an implicit alternative of driving. A better question is, "Do you like to fly when traveling short distances, or would you rather drive?" The first question is likely to yield a greater preference for flying.
Avoid Implicit Assumptions
Questions should not be worded so the answer depends on implicit assumptions about consequences. For example, "Are you in favor of a balanced budget?" has implicit assumptions about how it would be achieved. A better question is, "Are you in favor of a balanced budget if it would result in an increase in the personal income tax?"
Avoid Generalizations and Estimates
Questions should be specific. For example, "What is the annual per capita expenditure on cosmetics in your household?" is incorrect. Better questions are, "What is the monthly (or weekly) expenditure on cosmetics in your household?" and "How many members are there who use cosmetics in your household?" The researcher can then calculate the annual per capita expenditure.
π‘ Why this matters: Poorly worded questions lead to biased, confusing, or unusable data, wasting time and resources. Precise wording is essential for obtaining accurate and reliable responses.
Sequencing the Questions
After the wording of the questions is described, their sequence should be established.
π Guidelines for sequencing:
- Use a simple and interesting question in the beginning to capture the respondent's interest.
- General questions should be asked first and specific questions later. Moving from general to specific is called the Funnel Approach.
- Uninteresting and hard questions should be placed late in the sequence. Embarrassing, sensitive, dull, and complex questions, including personal ones about age and income, should be placed at the end.
There are three types of information gathered with a questionnaire: Basic information (relates directly to the research problem), Identification information (name, telephone, address), and Classification information (used to classify respondents for analysis). The general order is: Basic information first, followed by classification, and finally identification information.
Logical Order: The questionnaire should flow smoothly and logically from one topic to the next. When a new topic is introduced, a transaction statement should be given. Branching questions should be designed carefully. It is advised that the research questionnaire be divided into several parts for basic information and arranged logically.
β Key Takeaways
The most critical points from this lecture are: first, to master the distinction between unstructured (open-ended) and structured (multiple-choice, dichotomous) questions, knowing their respective advantages and disadvantages in terms of bias, cost, and depth of insight. Second, you must apply the rigorous guidelines for wording questions: define the issue, use ordinary and unambiguous words, and avoid leading questions, implicit alternatives, and implicit assumptions. Third, understand the proper sequencing of questions, which starts with simple, general questions and ends with difficult or sensitive ones (the "funnel approach"). Fourth, know the order of information types: basic first, then classification, then identification. Finally, remember that a well-designed questionnaire is a balance of clear structure, precise wording, and logical flow to minimize bias and maximize the quality of data collected.
π§ Quick Revision Questions
- What is the primary advantage of an unstructured question, and what is its major disadvantage?
- What is "order bias" in multiple-choice questions, and how can it be minimized?
- When designing a dichotomous question, under what circumstance should you include a neutral response alternative like "Don't know"?
- Why is the question "Do you like to fly when traveling short distances?" considered to have an implicit alternative, and what is the correct way to reword it?
- What is the "funnel approach" in questionnaire sequencing, and what is the recommended order for basic, classification, and identification information?
Here is the summary of Lecture 19, following the exact format and rules provided.
π Lecture 19 β Data collection and fieldwork
π Overview: This lecture details the critical steps in preparing a questionnaire for field use, focusing on how to introduce it via a covering letter, why and how to pre-test it, and the final process of pre-coding. Mastering these elements ensures higher response rates, data quality, and efficient data processing, which are essential for valid research outcomes.
ποΈ Topics Covered
The lecture begins by explaining the components and purpose of a covering letter for a questionnaire, including whether to disguise the sponsor. It then details the vital process of pre-testing the questionnaire to identify and fix errors, covering the purpose and methods like protocol analysis and debriefing. Finally, it concludes with the concept of pre-coding the questionnaire to assign numerical codes for efficient data entry.
π Lecture Summary
Covering Letter for Questionnaire
Introducing the questionnaire is an important step, usually done via a covering letter (or opening remarks in person). This introduction must contain: the name of the sponsor or survey team, why the survey is being conducted, how the respondents were selected, a request for response, and any incentive provided.
π‘ Why this matters: The covering letter is the first point of contact and significantly impacts a respondent's willingness to participate.
- Regarding the sponsor, the survey can be undisguised (researcher and sponsor identified) or disguised (sponsor concealed). Hiding the sponsor prevents competitor knowledge and prevents the sponsor's name from biasing responses.
- The purpose of the survey should be stated simply and clearly, sometimes even concealing the sponsor's name.
- Respondents must be told how they were selected (e.g., at random, judgmental, or by referral/snowballing) to address their curiosity.
- The letter must include a polite request for cooperation and participation.
- The final section may mention incentives, such as:
- Assurance of confidentiality and anonymity.
- A sample of a product.
- An offer to provide a copy of the research results.
- Monetary incentives (e.g., Rs. 100 for a complete questionnaire).
Pre-testing the Questionnaire
Before field use, a questionnaire must be pre-tested (used on a trial basis in a small pilot study) and revised to see how it performs under actual conditions. At least one pre-test and revision is vital.
The main purpose of pre-testing is to improve the questionnaire by checking for:
- Precise instructions for the respondent.
- Double-barreled questions (asking for two bits of information at once).
- Understanding wording, confusion, or double meanings.
- Leading or loading questions.
- Unfamiliar technical words or abbreviations.
- Whether the questions cover all conceivable answers.
- Adequate sequence and flow of questions.
- An appealing and motivating general appearance.
Method of Pre-testing the Questionnaire involves:
- Selecting 10-15 persons similar to the actual sample.
- Informing them of the pre-test's purpose and asking them to identify confusing elements.
- First, pre-testing via personal interview, then revising.
- Next, pre-testing the revised questionnaire in the final intended mode (mail, telephone).
- Protocol analysis: Respondents think aloud while answering; comments are recorded.
- Debriefing: After completion, respondents are told it was a pre-test and invited to comment on problems.
- Preparing dummy tables from pre-test responses to check for tabulation issues.
- Revising the questionnaire based on findings. If many changes are made, re-test with a new sample.
Pre-coding the Questionnaire
Pre-coding is the final task in questionnaire development, involving placing codes (numbers) on the questionnaire to facilitate data entry. The objective is to associate each response with a unique number or letter. Numbers are preferred as they are faster to keystroke into computers.
π Formula (Procedure): Assign a numeric code to each possible answer β this allows for fast and efficient data entry and computer processing.
π Example:
- Have you purchased Pizza A in the last fifteen days?
- Yes (1)
- No (2)
- The last time you bought Pizza A
- Was delivered to your house (1)
- You picked it yourself (2)
- Was delivered in your office (3)
- You ate the pizza at their restaurant. (4)
- What is your opinion about the taste of Pizza A?
- Excellent (1)
- Good (2)
- Satisfactory (3)
- Poor (4)
β Key Takeaways
The covering letter is crucial for securing respondent cooperation and must clearly state the surveyβs purpose, how the respondent was selected, provide an assurance of confidentiality, and offer an incentive. To ensure data quality, a questionnaire must be pre-tested, not just reviewed internally. Pre-testing should involve a small group of representative respondents and use methods like protocol analysis or debriefing to identify confusing questions, wording issues, and poor flow. The final step is pre-coding, where each possible answer is assigned a unique numerical value to make data entry and computer processing faster and more efficient. A poorly introduced, un-tested, or un-coded questionnaire will lead to biased data, low response rates, and logistical nightmares.
π§ Quick Revision Questions
- What are the five essential elements that must be included in a covering letter for a questionnaire?
- Explain the difference between a "disguised" and an "undisguised" survey, and give one reason for using the disguised approach.
- What is the main purpose of pre-testing a questionnaire, and what is a "double-barreled question" that pre-testing helps identify?
- Describe the difference between "protocol analysis" and "debriefing" as methods for pre-testing a questionnaire.
- What is "pre-coding" of a questionnaire, and why are numbers preferred over letters for this process?
π Lecture 20 β Data Analysis: Data preparation and data cleaning
π Overview: This lecture focuses on the final production of the questionnaire and provides a comprehensive final checklist for questionnaire design. It emphasizes the physical appearance, structuring, and pre-testing of questionnaires to ensure high response rates and data accuracy, which are critical for effective data analysis.
ποΈ Topics Covered
The lecture covers the final production of the questionnaire, detailing physical appearance considerations like paper quality, booklet format, question numbering, and layout. It then provides a comprehensive final checklist for questionnaire design, covering information specification, question content, response format, wording, sequencing, physical characteristics, and pre-testing procedures.
π Lecture Summary
Final Production of Questionnaire
After pre-testing and pre-coding, the questionnaire should be produced finally. The physical appearance of the questionnaire can affect the response rate and the accuracy of the answers. The following points should be kept in mind in the final production/reproduction of the questionnaire.
- Use good quality paper. A professional appearance is crucial; if the quality is poor, the respondent may think the research is unimportant.
- If the questionnaire runs to several pages, it should be in the form of a booklet instead of stapled sheets of paper.
- Questions should be numbered serially to help control in field operation, but avoid numbers if anonymity is an issue.
- Leave reasonable space for open-ended questions; too little space will result in short answers.
- Use "go to" instructions if the respondent must skip a question or section.
- Each question should fit on one page only; do not spread a question across two pages.
- Type and font should be large enough to be read clearly.
- Use vertical instead of horizontal response columns, as it is easier to read down the column.
- Color coding may be used, including different colors for different respondent groups.
- The questionnaire should appear short, but not at the expense of overcrowding the questions.
- Instructions and directions should be clear and placed close to the questions.
Final Checklist for Questionnaire Design
This section provides a detailed checklist organized into several key areas for designing a questionnaire.
Specify what information will be sought:
- Review the research problem, information needed, research questions, and hypotheses.
- Have a clear idea of your respondents.
- Determine the type of interviewing (personal, telephonic, or mailed questionnaire).
Determine the content of questions:
- Check if the question is necessary and how many are needed.
- Avoid double-barreled questions; ensure each question addresses only one issue.
- Use filter questions if the respondent is informed.
- Use aided recall if the respondent can remember.
- Provide necessary motivation if the respondent is willing to share information.
- Avoid threatening questions.
- Consider if the request appears legitimate and the level of effort required.
- If the information is sensitive, bring sensitive topics to the end, use third person technique, and use counter biasing statements.
π Definition β Double-barreled question: a question that asks about two or more issues at the same time, making it difficult for the respondent to answer accurately. π Definition β Filter question: an initial question that screens respondents to determine if they are qualified or informed enough to answer subsequent questions. π Definition β Aided recall: a technique used to help respondents remember information by providing cues or prompts. π Definition β Third person technique: a method of asking questions by referring to another person or group to reduce the respondent's discomfort with sensitive topics. π Definition β Counter biasing statement: a statement that acknowledges the sensitive nature of a question to make the respondent more comfortable and reduce bias.
Determine response Format/Structure:
- Decide on the appropriate structure: open-ended, structured, multiple choice, or dichotomous.
- Use open-ended questions in exploratory research; use structured questions whenever possible.
- In multiple choice questions, include all possible choices.
- Clearly indicate if items are to be ranked or only one item is to be selected.
- In dichotomous questions, include a neutral choice if many neutral opinions are expected.
π Definition β Structured question: a question with a predetermined set of response options, such as multiple choice or dichotomous. π Definition β Dichotomous question: a question with only two response options, such as "yes" or "no."
Determine the wording of the questions:
- Use simple language and words.
- Avoid ambiguous wording like "frequently" or "often."
- Avoid leading questions and loading questions.
- Avoid implicit assumptions and implicit alternatives.
- Use positive and negative statements.
- Don't make the respondent compute a lot.
π Definition β Leading question: a question that suggests or implies a particular answer, biasing the response. π Definition β Loading question: a question that is phrased to evoke a positive or negative emotional response. π Definition β Implicit assumption: an assumption that is not explicitly stated but is built into the question. π Definition β Implicit alternative: a question that presents only one option while implying there is a different one.
Determine the sequence of questions:
- Use a simple, non-threatening open-ended question at the beginning.
- Use the funnel approach: ask broad questions first, then narrow them down.
- Ask sensitive or hard questions late in the questionnaire.
- Follow chronological order when collecting historical data.
- Ask filter questions if needed.
- Complete questions about one topic before moving to the next.
- Use logical order in sequencing wherever possible.
π Definition β Funnel approach: a questioning technique that starts with broad, general questions and gradually narrows to more specific ones.
Determine physical characteristics of questionnaire:
- Ensure the questionnaire has a professional appearance.
- Start with an introduction (or separate introductory letter).
- Use good quality print and paper.
- Keep the questionnaire short but not overcrowded.
- Use a booklet form for many pages.
- List the organization's name on the first page unless there is a reason to disguise it.
- Number questions and pre-code answers for easy computer processing.
- Use "go to" instructions for skipping questions.
- Number the questionnaire in serial order.
- Use vertical response columns.
- Place directions or instructions close to the questions.
Pre-test and revise the questionnaire
Pre-testing is a crucial step to identify and fix problems before the main data collection.
- Do at least one pre-test; if time permits, do two or more.
- Pre-test first by personal interview, then by the method to be used ultimately.
- Pre-test all aspects: content, wording, form, sequence, and difficulty level.
- The pre-test sample may be small (15-30) for initial testing.
- Use protocol and debriefing procedures to identify problems.
- The pre-test sample should be drawn from the same population as the ultimate sample.
- Prepare dummy tables and analyze the data.
- Make revisions if necessary.
- If more than one pre-test, each should be done on a different sample of respondents.
π Definition β Protocol: a procedure where a respondent thinks aloud while completing the questionnaire, revealing any confusion or issues. π Definition β Debriefing procedure: a post-test interview where the respondent discusses their experience with the questionnaire to identify problems.
π‘ Why this matters: Proper pre-testing helps catch errors in question wording, skip patterns, and overall flow, saving time and resources while improving data quality.
β Key Takeaways
A student must remember that the physical appearance of a questionnaire directly impacts response rate and data accuracy, so use good quality paper, a booklet format if long, vertical response columns, and clear instructions. The final checklist covers key areas: specifying information needed, determining question content (avoiding double-barreled and threatening questions), response format, wording (using simple, unambiguous language), and question sequencing (using the funnel approach and placing sensitive questions last). Pre-coding answers and using βgo toβ instructions are essential for data processing ease. Finally, pre-testing the questionnaire on a small sample from the target population, using protocols and debriefing, and preparing dummy tables are critical steps to identify and fix problems before full deployment.
π§ Quick Revision Questions
- What are five key physical appearance features of a well-produced questionnaire that can improve response rate?
- What is a double-barreled question, and why should it be avoided in questionnaire design?
- What is the funnel approach to question sequencing, and where should sensitive questions be placed?
- What is the purpose of pre-testing a questionnaire, and what two procedures can be used to identify problems during pre-testing?
- Why should vertical response columns be used instead of horizontal ones in a questionnaire?
π Lecture 21 β Data Analysis: Data preparation and data cleaning...contd
π Overview: This lecture focuses on the critical phase after data collection β data preparation and cleaning. It covers the design and use of observational forms for recording behavioral data, and delves deeply into the critical issue of non-response error in surveys. The lecture explains the sources of non-response, particularly refusals and not-at-homes, and provides a comprehensive toolkit of strategies to improve response rates and reduce bias.
ποΈ Topics Covered
The lecture begins by detailing the structure and pre-testing of observation forms for recording behaviors like customer service interactions. It then introduces the concept of non-response error and its potential to bias survey results. The core of the lecture is dedicated to methods for improving response rates, organized into two main categories: reducing refusals (through prior notification, motivation, follow-ups, incentives, and good questionnaire design) and reducing not-at-homes (primarily through callbacks).
π Lecture Summary
Observation Form
An observational form is simpler to design than a questionnaire because it eliminates the question-asking process and helps control non-sampling error. The researcher must be very clear about what types of observations to make and how to measure them, using a form, a recording device (e.g., video), or both.
The form specifies details using the "Five Ws and How":
- Who to be observed? (Purchaser, browser, males, couples, anyone entering the store)
- What to be observed? (Brands purchased, quantity, inquiries, influence of others)
- When to observe? (Weekdays, weekends, date, hour of purchase)
- Where to observe? (Type of store, checkout counter, specific department)
- Why to observe? (Family purchase pattern, influence of price or brand name)
- Way to observe. (Participant observer, undisguised observer, hidden camera)
Observational forms should be simple to use and should allow the observer to record behavior in detail rather than summarizing it. Like a questionnaire, an observational form must be pre-tested.
π Example: A portion of an observational form to evaluate bank employee service includes a checklist for "Customer Relation Skills" with items like:
- Employee noticed and greeted immediately.
- Employee speaks pleasantly and smiles.
- Employee found your name promptly.
- Employee asked you to be seated.
- Employeeβs desk area was clean.
- Employee was helpful.
Each item is rated as Yes, No, or Does not apply.
Non Response Issue in Data Collection
Non-response error arises when potential respondents in the sample do not respond. This is a significant problem because if non-respondents differ from respondents on the characteristics of interest, the sample estimates will be seriously biased. While higher response rates generally imply lower non-response bias, they do not guarantee the respondents are representative of the original sample.
π‘ Why this matters: A low response rate increases the probability of error, so every effort must be made to improve it.
π Definition β Non-response error: The error that occurs when some potential respondents included in the sample do not respond, potentially biasing the results if they differ from those who did.
Improving the Response Rates
Improving response rates involves addressing two main sources of non-response: refusals and not-at-homes.
Refusals result from the unwillingness or inability of sampled people to participate. Strategies to reduce refusals include:
-
Prior Notification: Sending potential respondents a letter notifying them of the imminent survey. This reduces surprise and uncertainty, creating a more cooperative atmosphere.
-
Motivating the Respondents: Increasing interest and involvement.
- Foot-in-the-Door Strategy: The interviewer starts with a small request (e.g., "Will you take 5 minutes to answer 5 questions?"), to which most comply. This is followed by the large, critical request. The rationale is that initial compliance increases the chance of later compliance.
- Door-in-the-Face Strategy: The reverse strategy, starting with a large request that is expected to be refused, followed by the smaller, actual request. The lecture states that foot-in-the-door is more effective than door-in-the-face.
π Definition β Foot-in-the-Door Strategy: A compliance strategy that starts with a small request to increase the likelihood of agreement with a subsequent, larger request. π Formula: Small Request β (Compliance) β Critical Request π Example: Asking "Will you please take five minutes to answer five questions?" and, after agreement, asking the full survey questions.
-
Follow-up: Periodically contacting non-respondents. This is very effective in mail surveys, often requiring two or three additional mailings. Follow-ups can also be done by telephone, email, or personal contact.
-
Other Facilitators: Personalization (sending letters to specific individuals) is effective.
-
Incentives: Offering monetary or non-monetary incentives.
- Prepaid Incentive: Included with the survey. Has been shown to increase response rates more than promised incentives.
- Promised Incentive: Sent only to those who complete the survey.
- Non-monetary Incentives: Premiums and rewards like pens, pencils, or offers of survey results. The amount of incentive has a positive relationship with response rate, but the cost may outweigh the value of additional information.
-
Questionnaire Design and Administration: A well-designed questionnaire and skilled administration (in telephone/personal interviews) can decrease overall refusal rates. Interviewers should not accept a "no" response without an additional plea.
Not-at-Homes: This rate can be substantially reduced by employing a series of callbacks (periodic follow-up attempts to contact non-respondents).
π Definition β Not-at-Homes: The percentage of sampled individuals who are not available to participate at the time of initial contact. π Example: Making a call on a weekday evening and then calling back on a weekend to reach a working person.
β Key Takeaways
- Observation forms are a structured tool for recording behavior, requiring clear definition of who, what, when, where, why, and how. They must be simple, detailed, and pre-tested.
- Non-response error is a major source of bias in surveys, occurring when non-respondents differ from respondents on key variables.
- Refusals are a primary cause of non-response and can be reduced through strategies like prior notification, the foot-in-the-door technique, follow-ups, personalization, and prepaid incentives.
- Not-at-homes can be effectively minimized by employing multiple callbacks at different times.
- The goal of all these techniques is to increase the response rate, thereby decreasing the probability and magnitude of non-response bias and improving the representativeness of the sample.
π§ Quick Revision Questions
- What are the six key elements an observation form must specify (the "Five Ws and How")?
- What is the difference between the foot-in-the-door and door-in-the-face compliance strategies for motivating respondents?
- Why is non-response error considered one of the most significant problems in survey research?
- Compare and contrast a prepaid incentive with a promised incentive. Which is generally more effective at increasing response rates?
- What is the primary method recommended for reducing the "not-at-homes" category of non-response?
π Lecture 22 β Descriptive and Inferential Statistics
π Overview: This lecture covers the structure and purpose of a research proposal, which is a written plan for a research project that serves both as a road map for the researcher and a persuasive tool for selling their services. It details the key components that must be included in a professional research proposal, from defining the problem to planning fieldwork. Understanding how to craft a comprehensive research proposal is essential for any marketing researcher seeking to secure contracts and conduct successful projects.
ποΈ Topics Covered
The lecture begins by defining a research proposal and explaining its dual purpose as a planning document and a persuasive technique. It then provides a detailed breakdown of the parts of a research proposal, starting with the research topic and executive summary, followed by the introduction, problem definition, research objectives, literature review, and research design. The final section of the proposal outline covers fieldwork, including data collection methods, personnel management, and quality control.
π Lecture Summary
Research Proposal
A research proposal is a written document that presents a plan for a project. Its primary purpose is to convince the client that the researcher is capable of successfully conducting the proposed research project. It serves as a planned document and a road map for the researcher. π‘ Why this matters: For professional researchers, the proposal is a key marketing tool for selling their services to clients.
Parts of a Research Proposal
- Research Topic: On the top of the proposal, write the research topic.
- Executive Summary/Overview/Abstract: A brief summary of what is given in the proposal. This is the part that busy executives will read.
- Introduction/Background: Before writing this, the researcher should meet with the manager to understand:
- What is the background to the research?
- What is the context in terms of the management problem?
- What is the environment?
- This part shows the researcher's understanding of the problem.
- Define the Problem: The researcher extracts the research problem from the introduction. The research problem is defined in one or two sentences and describes the main purpose, scope, and limits of the research.
- Research Objectives: Specific objectives or information needs must be specified clearly and concisely. These objectives should be SMART (specific, measurable, achievable, realistic, and timely).
- Literature Review: A brief review of related literature is given, referencing all books, journals, websites, and newspapers. A review of literature helps in developing:
- Theoretical framework/models
- Hypothesis
- Questionnaire
- State Research Design: This section describes the research design (descriptive, causal, or exploratory) in detail and must include:
- Nature of Data: The data is either primary or secondary.
- Sources of Data:
- Primary data sources include people, stores, brands, customers, and managers.
- Secondary data sources include books, government reports, UNO reports, ministry of agriculture/industry reports, or independent researchers, consultants, or syndicated data. They may be internal and external.
- The list of sources must be provided in the research methodology.
- Population and Sample Size:
- Population is the aggregate of all the units which have the data for the research.
- The researcher must indicate the size of the sample and the sampling techniques for selecting it.
- Instrument of Data Collection: The researcher defines which instrument will be used (questionnaire, observation form, or experimentation).
- In the case of causal research, the researcher must define the control and experimental group.
π Definition β SMART: An acronym for objectives that are specific, measurable, achievable, realistic, and timely.
Field Work
This section describes fieldwork activities, including:
- How the data will be collected.
- How many people/field force will be required.
- How they will be recruited.
- How they will be trained and motivated.
- How they will be supervised.
- How the quality of research will be controlled.
- The researcher specifies the control mechanism to ensure data quality. Supervision of the field force is very important to maintain the quality of research.
β Key Takeaways
A research proposal is a critical document that serves as both a planning roadmap for the researcher and a persuasive sales tool for the client. It must begin with a clear problem definition and set SMART objectives. The core of the proposal is a detailed research design that specifies the nature and sources of data, the target population, the sample size, and the data collection instrument. A comprehensive fieldwork plan is equally essential, covering everything from personnel recruitment and training to supervision and quality control. Successfully incorporating all these elements demonstrates the researcher's competence and ability to deliver a quality project.
π§ Quick Revision Questions
- What are the two main purposes of a research proposal?
- What does the acronym SMART stand for in the context of research objectives?
- What are the two main categories for the "Nature of Data" in a research design?
- List three items that a "Sources of Data" section can include for secondary data.
- What are the key activities that must be planned for in the "Field Work" section of a proposal?