Research basics; sample and measurments) Students.pdf

1 Research basics; Samples and measurments Dr Sara A. Zaher MSc, PhD, pediatrics & critical care Nutrition Assistant Professor Clinical Nutrition Department Taibah University  Outlines...

1 Research basics; Samples and measurments Dr Sara A. Zaher MSc, PhD, pediatrics & critical care Nutrition Assistant Professor Clinical Nutrition Department Taibah University  Outlines 2 1. Measurment error. o Random error. o Systematic error. 2. Sampling ( recruitment) methods. o Random. o None Random 3. Tools for data collection. o Questionnaires. o Interview. o Focus group. o Observation methods. o Documents analysis.  Learning outcomes To understand and distinguish between different sampling methods. To be able to choose the optimal tools for data collection based on the variables intended to be measured. Measurment Error 4  Types of Errors 1. Random error is a wrong result due to chance 2. Systematic error is a wrong result due to bias  Sources of Error  Observer variability. ( data collector)  Instrument variability.  Subject variability.  How do we minimize measurment error? 7 1. Increasing the percision. (Reliability) 2. Improve the accuracy. (Validity)  What is accuracy and precision 8  Accuracy and precision are alike only in the fact that they both refer to the quality of measurement, but they are very different indicators of measurement.  Precision is the degree to which an instrument or process will repeat the same value. (Precise values differ from each other because of random error).  Accuracy is the degree of closeness to true value. (The greater the error, the less accurate the variable)  Precision Example: Who is more precise when measuring the same 21.0 cm book four times? Hind 17.0 cm, 19.0 cm, 20.0 cm, 21.0 cm Marya 15.5 cm, 15.0 cm, 15.2 cm, 15.3 cm  Accuracy Example: Who is more accurate when measuring the same 21.0 cm book four times? Hind 17.0 cm, 19.0 cm, 20.0 cm, 21.0 cm Marya 15.5 cm, 15.0 cm, 15.2 cm, 15.3 cm Sampling Methods 12  Basic Concept  Population & sample: o Population: total collection of entities of interest, usually defined by the study designer. Example: Population Sample Research question: Is smoking a risk factor for lung cancer in Saudi Arabia?  Population: All lung cancer patients and smokers in Saudi Arabia.  Sample: A representative part of the population. e.g Lung cancer study – a sample of 10,000 smokers and 500 lung cancer patients.  Sampling method is the way of selecting the subjects in the research (Recruitment)  There are basically two main types of sampling:  Probability (random).  Non-probability (non-random). Probability Sampling 16  1. Simple Random Sampling  Each sample should have equal and independent chance to be selected.  This is the simplest form of sampling and the ideal method.  Could be done by tossing a coin, throwing a dice or blindfolded method. Example: 20 subjects are available and the target is to recruit 4 of them. What to do? Write each of their names on a piece of small paper randomly draw four of those rolled papers.  2.Systematic Random Sampling Steps: 1. Organize the individual of the population in any form of order (e.g. alphabetically or by their address …) this is called sampling frame 2. Start from a random point (e.g. from the second individual). 3. Set a regular interval (e.g. every 3rd will be selected).  3.Cluster Random Sampling  In cluster random sampling subjects were distributed, ideally in homogenous groups that we called cluster.  To represent a state that contains 4 districts, is it possible to simply select 1 out of the 4 district randomly.  That one district selected shall represent the entire state.  Selecting 1 district may require a logistic reason e.g. Cost effective focus on one district rather than concentrating on getting all 4 districts.  Cont’ What if the clusters are not exactly homogenous? This technique will introduce bias in the measurement of variance (design effect). Give examples of heterogeneous clusters? Studying the population of the world based on focusing on one country  4. Stratified Random Sampling  Is a method of sampling that involves the division of a population into smaller subgroups known as strata.  The strata are determined based on certain characteristics such as sex, age groups and location. Example:  Selecting 4 subject out of 20 (male and female)  2 samples from each sex shall be randomly selected Non-probability Sampling 23  1. Convenience (accidental) sample  Subjects on the time of selection are selected because they were available. Example: anyone admitted to the ER from 4:00-7:00 pm in a particular day(s)  2. Purposive sample  The idea is to pick out the sample in relation to criterion which are considered important for the particular study.  This method is appropriate when the study places special emphasis upon the control of certain specific variables.  26 Example:  Targeting university students to collect their inputs about the education system and their choice of subject using a Student Feedback Survey.  3. Snowball sample  This is a sampling technique, in which existing subjects provide referrals to recruit samples required for a research study.  This sampling method involves a primary data source nominating other potential data sources that will be able to participate in the research studies.  Cont’ 28 Example:  A research is to be conducted on vegetarians. At the first stage, we may pick up a few people who are known to us or can be identified to be vegetarians. At the time of interviewing them, we may obtain the names of other vegetarian people known to the first stage subjects.  Thus, the subjects go on serving an informant for the identification of more subjects and the sample goes on increasing. Tools for Data Collection 1. Questionnaires. 2. Interview. 3. Focus group. 4. Observation methods. 5. Documents analysis. 6. Others.  Introduction 32 The researcher needs to ensure that the proposed instruments are as valid before the study begins (based on literature). The quality of the results depends on the quality and appropriateness of the method.  Quesionnaires, interviews, focus groups 33 These three methods of data collection involve asking questions to participants.  Types of questions Types of questions Open ended Close ended  1.Open-ended  Open-ended questions are particularly useful when it is important to hear what respondents have to say  Example: What habits do you believe increase a person’s chance of having a stroke?  Open-ended questions may used in exploratory phases as they help the researcher understand a concept as respondents express it.  Phrases and words used by respondents can form the basis for closed-ended questions.  Advantages and disadvantage Advantages Disadvantages Free to answer with fewer limits imposed. Time consuming (To code and analyze the responses) Report more information  2.Closed-ended questions 2.1 Questions. 2.2 Visual analogue scale (VAS). 2.3 Likert-scales.  2.1 Questions  Questions that provide a list of possible alternatives from which the respondent may choose. Advantages Disadvantages Less time (quicker) Restricted answers Easier to answer The set of answers may not be exhaustive Answers are easier to tabulate and analyse.  2.2 Visual analogue scale (VAS)  Another option for recording answers to closed-ended questions using lines or other drawings.  Each end describe the most extreme values.  Example: Mark the place on this line that best describes the severity of the pain:  2.3 Likert-scales: Commonly used to quantify attitudes, and behaviours. Provide respondents with a list of statements to select best rank or degree. Likert scales can measure variations such as frequency, quality, importance, and likelihood, etc. Each response is assigned a number of points.  41 Example: A questionnaire to measure the strength of a person’s opinion about a diet high in fruits and vegetables improves health: 42  1. Questionairs 43 Questionnaires are instruments used to gather data from respondents, that they fill out by themselves. Questions usually close ended. Inexpensive approaches to collect health survey information. Questionnaires can be completed by:  Subjects in person.  E-mail, or through a website.  Handheld electronic devices.  Designing questionnaires 1. Wording and clarity:  Make questions as clear, short and specific as possible  Provide an example.  The question should match the options for the answer Example: Which option is correct? o Have you had pain in the last week? (Never /seldom /often /very often). o How often have you had pain in the last week? (Yes / No) Which option is correct? Discuss o I am sometimes depressed (agree/ disagree) o I am sometimes depressed (never/ sometimes/ often/ always)  45  Avoid double-barrelled questions: (Each question should contain only one concept) Example: A question was designed to evaluate caffeine intake: How many cups of coffee or tea do you drink during a day?  It is known that coffee contains much more caffeine than tea  It is better to break it into two separate questions  46 2. Simplicity:  Use simple, common words and grammar  Drugs you can buy without a prescription from a doctor. or  Over-the-counter medications?  Cont’ 47 3. Organization and formatting:  Questions concerning major subject areas should be grouped together (headings or short descriptive statements).  Highly sensitive questions are often placed at the end.  48 4. Neutrality and tone: a. Avoid “loaded” questions.  How often did you smoke too much cigarettes? Or  How often did you smoke more than five cigarettes in one day? b. Collecting information about sensitive areas e.g. income is difficult. (grouping)  Cont’ c. Avoid setting a tone that permits the respondent to admit to behaviours that may be considered undesirable. Example: People sometimes forget to take medications their doctor prescribes. Does that ever happen to you?  2. Interviews 50 Interviews are a method of data collection that involves two or more people exchanging information through a series of questions and answers.  Cont’ 51 Examples where interview approach is useful:  Researcher wish to to ask a questions that require lengthy explanation or want to gather very detailed information, e.g. Asking about the current eating behavior.  Researcher anticipate wanting to ask respondents follow-up questions based on their responses.  Researcher is studying a complex or potentially confusing topic to respondents.  Researcher is studying processes, such as how people make decisions.  Cont’ 52  Questions in interviews usually open ended.  The researcher need to take the following points into account when asking questions: 1. Avoid Pitfalls: (Closed Questions, Insignificant Questions, Hypothetical Questions, Leading Questions, Irrelevant Questions, Vague Questions, Impossible Questions) 2. Hidden assumptions: assumptions that may not apply to participants in the study should be removed. Example: I could not feel better even with help from my family  This assumes that respondents have families and ask for emotional support  Cont’ 53 Pitfall Example Revision Principle “Is it significant that Jesus is Ask questions that allow for 1. Closed Questions “To whom is Jesus talking?” talking to a Samaritan woman? various answers. Why or why not?” “What do you notice about the “What do you observe about the setting that might change or Ask questions that convey 2. Insignificant Questions setting?” influence the way you understand significance. this account?” “What are some ways the Ask questions that focus on the “What would you do if you were 3. Hypothetical Questions Samaritan woman could have text rather than hypothetical the Samaritan woman?” responded to Jesus?” scenarios. “What do you think is Jesus’ Ask questions that allow group “How does Jesus prove He is the main objective in talking to the 4. Leading Questions members to discover meaning for Messiah?” Samaritan woman? What clues do themselves. you have?” “How old is the Samaritan “Do you think it matters how old Ask questions that relate directly 5. Irrelevant Questions woman?” the Samaritan woman is?” to the main points of the passage. “Do you notice differences in “What is the difference between Ask questions that are clear and 6. Vague Questions how Jesus and the Samaritan Jesus and the Samaritan woman?” specific. woman talk to each other?” Ask questions that can be “Why didn’t Jesus perform a “Was there anything curious answered by someone with 7. Impossible Questions miracle?” about Jesus’ behavior?” general—not specialized— knowledge.  Comparison between questionnaires and interveiw approaches Questionnaires Interviews Advantages Disadvantages Advantages Disadvantages Generally a more Not suitable for Appropriate for More costly & time- efficient difficult question or complicated questions consuming need explanation (require explanation or guidance) Questionnaires are less Questionnaires some Completing the Standardizing could be expensive than times are not questions a problem interviews (require less completed. research staff time) Easily standardized Privacy  3. Focus group 55  A focus group is a research method that brings together a small group of people to answer questions in a moderated setting.  The group is chosen due to predefined demographic traits, and the questions are designed to shed light on a topic of interest. Example: A focus group discussion of pregnant women to understand the psychological changes during the first trimester. Observation & decument analysis 56  4. observation 57 Observation is way of gathering data by watching behavior, events, or noting physical characteristics in their natural setting. Example: Observing the weight changes in critically ill patients staying in ICU for prolonged period.  5. Decument analysis 58 Document analysis is a method of data collection that require access to essential existing sources such as documents and records. Requirements:  Significant records and documents to extract data from.  Gaining access or permission to these documents. Example: Extracting anthropometric and enteral feeding data from the hospital’s records or electronic system to assess the incidence of underfeeding among critically ill children. 59 Questions