NUT602: Research Methods in Nutrition and Food Science Data Collection PDF

NUT602: Research Methods in Fall Nutrition and Food Science 2024 Data Collection  Measurement 2 Type of outcomes influences study design and determines the analytical tool to be used to study the relationship of interest ...

NUT602: Research Methods in Fall Nutrition and Food Science 2024 Data Collection  Measurement 2 Type of outcomes influences study design and determines the analytical tool to be used to study the relationship of interest  Validity 3 Reliability Levels Dose- normalize in response remission relationship Measurement: Desired characteristics Changes of a surrogate Surrogate accurately in response to Cut-off point for normality treatment High Availability sensitivity Affordability Specificity Data Collection 4 ▪ The process by which the researcher collects the information needed to answer the research problem ▪ The researcher must decide ▪ Which data to collect? ▪ How to collect the data? ▪ Who will collect the data? ▪ When to collect the data? ▪ We use research instruments (tools) to collect data ▪ You either construct a research instrument or used one already have been used ▪ The tool should be pre-tested on a pilot group Main Stages in Data Collection 5 ▪ Stage 1: Permission to proceed ▪ Ethical approval ▪ Informed consent ▪ Stage 2: Data collection ▪ Logistics ▪ Quality control (data need to be valid and reliable) ▪ Stage 3: Data handling ▪ Proper coding and numbering ▪ Proper storage  6 Possible Sources of Bias in Data Collection ▪ Not following sampling procedures planned ▪ Bias in measurements due to ▪ Changes in participant behavior affected by study itself (blood pressure is high due to measurement) ▪ Use of unstandardized equipment ▪ Observer bias/variability ▪ Changes in measurement criteria or ways to record data Methods of Data Collection 7 7 Survey 8 ▪ Non-experimental research method used for collecting data from a pre- defined group of respondents to gain information on topics of interest ▪ Advantages ▪ Good for gathering descriptive data ▪ Can cover a wide range of topics ▪ Are relatively inexpensive to use ▪ Can be analyzed using a variety of existing software ▪ Disadvantages ▪ Self-report may lead to biased reporting ▪ Data may provide a general picture but lack depth ▪ May not provide adequate information on context  Observation 9 ▪ Methods by which an individual or individuals gather firsthand data on programs, processes, or behaviors being studied ▪ Advantages ▪ Provide direct information about behavior of individuals and groups ▪ Permit evaluator to enter into and understand situation/context ▪ Provide good opportunities for identifying unanticipated outcomes ▪ Exist in natural, unstructured, and flexible setting ▪ Disadvantages ▪ Expensive and time consuming ▪ Need well-qualified, highly trained observers (content experts) ▪ May affect behavior of participants ▪ Selective perception of observer may distort data ▪ Behavior or set of behaviors observed may be atypical  10 Research Instruments ▪ Ways of searching for research instruments 1. Scientific publications (to learn what kind of instruments are being used for similar studies, their format, style and how they were used by the authors) 2. Books (that provide a description or an actual copy of various instruments) 3. Other researchers (who may know of certain tools they have developed for themselves, or may used tools developed by others) 4. Combine or adapt one or more tools (used by other researchers) 5. Develop own instrument  11 Instrument Development ▪ Requires a high degree of research expertise ▪ The instrument must be ▪ Reliable (consistency of a measure) ▪ Valid (accuracy of a measure) ▪ Suitable for its function ▪ Content appropriate to test the hypothesis or answer the question being studied ▪ Free of build-in clues ▪ Types of Instruments 1. Questionnaire 2. Checklist 3. Records …  9 STEPS FOR THE DEVELOPMENT OF A QUESTIONNAIRE 12 Steps Examples Describe the information required Dietary behavior, PA, food insecurity Describe the target participants Pregnant women, children, older adults Select the research methodology appropriate for FFQ, 24-hr dietary recall, diet history the target population Decide on question content Nutrient intake, PA levels, demographics Develop the question wording Avoid technical terms, be clear and concise, measure one thing per question, avoid leading questions Put questions into a meaningful order and format Administer in person, via email or by telephone Check the length of the questionnaire Keep the questionnaire short Pretest the questionnaire Pilot test for feedbacks and modifications Develop the final survey form Get permission from participants. If online, check how it looks and readability  13 Questionnaire ▪ An instrument used for collecting systematic, standardized information that can be expressed numerically or in short answers ▪ A series of questions, can be gathered either by oral interview or by written questionnaire ▪ Most common type of research instrument Questionnaires are useful when ▪ Data needs to be collected from large numbers of people ▪ Identical data needs to be collected at more than one point in time ▪ Standardized instruments exist or can be developed Questionnaire 14 Interview-Completed ▪ Interviewer asks questions (e.g., in-person, telephone…) ▪ Interviewer records responses Self-Administered ▪ Respondent receives survey (e.g., mail, email, Internet…) ▪ Respondent completes on his or her own Questionnaires – Interviewer-Completed 15 Advantages Disadvantages Greater instrument High cost complexity Interviewer bias Diverse populations Subject reluctance reached Safety concerns High response rate Anonymity not Opportunity to clarify for guaranteed respondents Scheduling Questionnaires – Self-Administered Advantages Disadvantages Low cost Reading ability required Rapid response Less control of data Validity of instructions Few open-ended Respondent can complete questions at their convenience Lower response rates Respondent motivation  9 STEPS FOR THE DEVELOPMENT OF A QUESTIONNAIRE 16 Steps Examples Describe the information required Dietary behavior, physical activity, food insecurity Describe the target participants Pregnant women, children, older adults Select the assessment tool FFQ, 24-hr dietary recall, diet history Decide on question content Nutrient intake, physical activity levels, demographics Develop the question wording Avoid technical terms, be clear and concise, measure one thing per question, avoid leading questions Put questions into a meaningful order and format Administer in person, via email or by telephone Check the length of the questionnaire Keep the questionnaire short Pretest the questionnaire Pilot test for feedbacks and modifications Develop the final survey form If online, check readability Attributes of a Good Questionnaire 17 ▪ Clarity of language ▪ Singleness of objective ▪ One-to-one correspondence (to study objectives): every item should relate to the topic under study ▪ Correct grammar, spelling, and construction ▪ Avoids double-negative questions ▪ Well stated title and statement of purpose ▪ Assures the respondents about the confidentiality of responses ▪ Has clear direction and design corresponds to easy data tabulation  Be Careful … 18 ▪ The questionnaire must contain simple but straight forward directions for the respondents so that they may not feel any difficulty in answering the questions ▪ Questions should be simple and constructed in a logical format (appropriate sequence) ▪ The layout and spacing are extremely important. If your questionnaire looks cluttered, respondents will be less likely to fill it in ▪ If you are constructing a combined questionnaire, keep open-ended questions for the end as, once someone has spent time completing the rest of the questionnaire, they are more likely to continue with those questions which take a little more effort to complete ▪ Place the delicate questions at the end of the questionnaire ▪ The researcher must invariably re-examine, and in case of need may revise the rough draft for a better one The Pre-test 19 ▪ To improve the quality of the questionnaire, it is desirable to ▪ Have the questionnaire re-read by experts ▪ Put the questionnaire to the test of the field ▪ A pilot study should be undertaken for pre-testing the questionnaire ▪ Administer the questionnaires to the same type of people who will be taking part in the main survey (sample of 10 to 50 people representative of the diversity of the studied population) ▪ Make sure they know it is a pilot test and ask them to forward any comments they may have about the length, structure and wording of the questionnaire ▪ Make sure the conditions are similar to the planned administration (identical place of investigation, etc.) ▪ The questionnaire may be edited based on the results of the pilot study  20 Realization of the Pretest ▪ In two phases 1. Observation phase of the questionnaire: list of questions, respondents' reactions, duration, refusals to answer, etc. 2. Phase of discussions with the “pre-testers”: overall assessment of the questionnaire, understanding of the questions, difficulties encountered, suggestions for improving questions and vocabulary, etc.  Let’s Practice 21 ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 1. Do you exercise? ▪ Never ▪ Rarely ▪ Frequently ▪ Sometimes The problem is in the categories supplied for the answer Everybody has a different idea as to what words such as ‘sometimes’ and ‘frequently’ mean. Instead, give specific time frames such as ‘twice a week’ or ‘once a week’ Also, the order of answers should follow a logical sequence –in this example, they do not Let’s Practice 22 ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 2. When your child is sick, do you go to the doctor? ▪ Yes ▪ No This seemingly simple question is poorly worded. The word "sick" is imprecise. For some parents, if the child has a cold, he is sick while for others he must have more fever  23 Let’s Practice ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 3. Does your child have asthma or allergies? ▪ Yes ▪ No This question should be separated into 2 different questions  Let’s Practice 24 ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 4. Do you agree that public health professionals should earn more money than they currently earn? This is a very leading question. It assumes that public health professionals should earn more money, and assumes that the respondent knows about the issue and is able to answer A revised version of this question could be: Do you believe that public health professionals’ salaries are a lower than they should be, higher than they should be, or about right? Public health worker salaries are lower than they should be Public health worker salaries are higher than they should be Public health worker salaries are about right Does not know/No opinion  25 Let’s Practice ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 5. What is wrong with breastfeeding in public places and what can we do about it? This question is double-barreled, leading and ambiguous It asks two questions in one and so needs to be split up The word ‘wrong’ is emotive and suggests there is something not normal about the breastfeeding in public places Let’s Practice 26 ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 6. How many books have you read in the last year? ▪ None ▪ 1-10 This question may contain prestige bias –would people be more likely to say they have read plenty of books when they might ▪ 10-20 not have read any? ▪ 20-30 Also, the categories for the answers need modification: which ▪ Over 30 box would you tick for someone who answered 10 or 20? Let’s Practice 27 ▪ Read the following question and decide what is wrong with it ▪ Think about how you might overcome the problems you have identified 7. Are you against abortion? ▪ Yes ▪ No If necessary, delicate questions in an indirect way - Direct mode: Are you for or against abortion?  For  Against - Indirect mode: Here are some affirmations, we ask you to express your degree of agreement: A woman must have the opportunity to have an abortion?  Agree  Neutral  Disagree  28 Coding of the Questionnaire ▪ To facilitate the continuation of the research, it is relevant to anticipate in the construction of the questionnaire the steps which follow the collection of the data which are the encoding (introduction of the questionnaire on computer and the analysis of the encoded data) ▪ Coding of the variables studied = assignment of a specific value to each of the ways of responding to a variable ▪ Example: Gender: Men=1 and women=2 ▪ Pre-code the questionnaire (show the coding on the questionnaire) ▪ Choose coherent codes (weight coded in kg) ▪ For complex variables formulated in an open form (the profession, for example), the codification may be subsequent to the collection of the data Checklist 29 ▪ Form used for quickly and easily recording data ▪ Items that comprise several questions on a topic and require the same response format 29 Records 30 ▪ A record refer to all numbers and statistics that institutions, organizations and people keep as record of their activities (i.e. hospital/clinic record ….) ▪ Advantages ▪ Unbiased ▪ Cover a long period of time ▪ Inexpensive ▪ Disadvantages ▪ If the record is incomplete, there is no way it can be completed ▪ No one can be sure of the conditions under which the records were collected ▪ No assurance of the accuracy of the records  RECAP: How to design a study 31 Study aims, background, rationale Endpoints or outcome variables, other assessments Specific variables, how measured, specific & sensitive to changes expected Reliability & validity of measure Inclusion/Exclusion criteria Accrual plan and preparatory tasks Timeline for overall study Timeline for individual study participant Treatments? Participant implications? Product, dose, quality, administration, and reproducibility of interventions (including training)  32 Take-home message