bioepi lab - data collection, processing, and presentation

• Data Collection procedure can use questionnaires, interview, review of records, and observation
• Questionnaire is a series of questions designed to collect information. It is the most common type of instrument used and is typically filled out by participants
• Questions can be open-ended or closed-ended
• Open-ended questions can elicit more detailed responses and the responses require more effort to encode for data analysis
• Closed-ended questions are easy to administer, uniform and pre-coded, and can be encoded and analysed in a short time
• Interview is a verbal communication between the researcher and the participant, during which information is collected
• An Unstructured interview is more conversational and it allows flexibility in questioning
• A structured interview operates within a formal interview schedule and order of questions are designed prior to the interview
• Sampling in interview includes random sampling and non-representative sampling
• In non-representative sampling, there's Key Informant interview where the interviewer conduct a one-on-one interview, and Focus Group Discussion where a small group of people is interviewed at the same time to informally discuss specific topics under the guidance of a moderator
• Review of Records is the collection of data from existing records using an abstraction form. Few example sources are hospital or facility records, computer data bases, government reports, census data, and more
• Observation is where behaviors are observed and measured using observation checklists or rating scales
• Observation requires us to develop standard protocol for data collection, blinding, train observers, and calibrate equipment
• The advantages of questionnaires are that it can assess a large group quickly and it's easy to analyze if constructed correctly. Its disadvantages on the other hand, requires "good" language, social desirability bias, and not very good in getting in-depth information
• The advantage of an interview is that you can know what people think, believe, or perceive. However, it can recall bias and a social desirability bias
• The advantages of records are its relatively inexpensive and faster than colleting the original data again. However, coding errors (missing or incomplete data) may occur, data may not be exactly what is needed, there's difficulty in getting access, and it needs to verify the validity and reliability of data
• The advantages of observation is that it's also relatively inexpensive and it collects data on actual vs self-reported behavior or perceptions. However, hawthorne effect and observer bias can occur. and can be labor intensive
• Data Processing is a systematic procedure to ensure that the information/data gathered are complete, consistent and suitable for data analysis
• Blinding or masking is when the participant/s does not know that they are under observation
• Hawthorne Effect is an alteration or awareness of the patient if they find out their behavior is under observation
• Data coding is transforming collected information or observation into numbers which can be more easily encoded, counted, and tabulated
• The purpose of data coding is that it allows rapid storage of data, minimizes errors in encoding data, sometimes necessary so that the statistical software can perform various analyses on the data
• Number of codes must be kept to minimum, preferably less than 8
• Data Coding should be exhaustive and mutually exclusive
• In data coding, one should adopt coding convention for questions with similar answers
• A coding manual consists of indicating variable name, variable description, and codes instruction
• Data Encoding is entering data in a spreadsheet. It uses computer programs for encoding
• Data Editing is the inspection and correction of any errors or inconsistencies in the information collected
• The purpose of Data Editing is to make changes or corrections as early as possible, to ensure completeness, consistency, and legibility of data entries, and to prepare the data for analysis
• Data Analysis is the process of evaluating data using analytical or statistical tools to discover useful information
• In Data Analysis, one should consider the objective of the analysis, level of measurement of the variable, and study design
• Relationship Test is a test for the significance of the relationship of variables (e.g. Pearson's Correlation Coefficient)
• Difference Test is the test for the significant differences in the group that is being compared to
• Parametric Test is making assumptions about the parameters of the population distributions from which the data are drawn
• Non-parametric Test is making NO ASSUMPTIONS about the parameters of the population distributions from which the data are drawn
• Study Design dictates the number of groups to be compared and dictates whether the samples are independent or related
• The purpose of data presentation is to display data clearly and effectively, summarize large quantities of data to the reader, and facilitate analysis of trends, comparisons, or relationships between variables
• Tabular Presentation is a compact way of presenting data in columns and rows
• The parts of a Table includes the Table Number, Title, Box Head (column heading), Stubs (row heading), Body, Footnotes, and Source of Data
• Table Number refers to the relative position of the table, placed on the same line as the title but separated by period, and uses Arabic numerals