All Weeks Data Management for Clinical Research Coursera Quiz Answers
This course presents critical concepts and practical methods to support the planning, collection, storage, and dissemination of data in clinical research.
Understanding and implementing solid data management principles is critical for any scientific domain. Regardless of your current (or anticipated) role in the research enterprise, strong working knowledge and skillset in data management principles and practice will increase your productivity and improve your science.
Our goal is to use these modules to help you learn and practice this skill set. This course assumes very little current knowledge of technology other than how to operate a web browser. We will focus on practical lessons, short quizzes, and hands-on exercises as we explore together best practices for data management.
Data Management for Clinical Research Coursera Quiz Answers
Week 01: Data Management for Clinical Research
Q1. Which of these examples describes a study eligible for a human subjects research exemption?
- A vaccine trial in a population of people in which there is an outbreak of a new disease.
- An evaluation of the learning effectiveness of a biology course for teenagers.
- A longitudinal study of health outcomes in participants consuming a vegetarian diet.
- A survey of patients’ attitudes and behaviors as they utilize services offered by a large hospital system, with questions about contact information for follow-up.
Q2. Which of the following is an example of an observational study?
- You assign children who have sore throat and fever into two different groups that get two types of antibiotics and you observe the duration of their sickness.
- You measure whether patients with diabetes who tell you that they are trying to lose weight can control their blood sugar better than patients who are not trying.
- You randomly pick patients with diabetes and offer them free subscription to a gym and nutrition service for losing weight and you observe whether their blood sugar is controlled better than the patients that were not randomly picked.
- All of the above.
Q3. The best approach for determining the variable type to store a physiological measurement of blood LDL cholesterol level in your study database is:
- You record the numerical value of the blood LDL cholesterol measurement and you determine whether it was high, low, or moderate when you are analyzing the data.
- You create an ordinal variable with the following codes: “1” for “ideal” LDL levels below 100 mg/dL; “2” for “normal” levels between 100-129 mg/dL; “3” for borderline” high between 130-159 mg/dL; and “4” for “high” if the LDL level is higher than 160 mg/dL.
- You need to minimize the size of your database. You can save disk space by creating a True / False variable for storing whether the LDL level is above 160 mg/dL instead of saving the actual numerical value of the LDL measurement.
- You don’t record the measurement itself, but ask an expert physician whether the patient had high or low LDL cholesterol level and you save their response into the database.
Q4. You should plan to collect data on all confounding variables, but sometimes research reveals new confounders during the course of the study. Which of the following is a possible strategy for collecting unforeseen confounding effects?
- Collecting samples for DNA testing to account for unforeseen knowledge about genetic associations that may be revealed in the future.
- Carefully studying the regulatory requirements and ethical considerations that underlie your study.
- Baseline Body Mass Index (BMI) is always a confounding variable and needs to be measured in every study.
- Documenting your variables in a codebook for use during data analysis.
Q5. Which of the following practices for handling missing or incomplete data is NOT recommended?
- Allow for uncertainty in recording dates (e.g., providing a way to record dates even if only the year or month are known), especially when asking participants to recall or estimate dates.
- Use a physically possible value as the code for missed or incomplete data (e.g., if age unknown, enter ’99’).
- Designate specific codes within the fields for representing missing, unknown, or incomplete data.
- All of the above are good strategies for handling missing or incomplete data.
Q6. Pick the sentence that makes most sense in the context of clinical research data management:
- You do not need to worry about data standards or sharing your data dictionary if you are collecting HIPAA identifiers since you cannot share that data anyway.
- Email is generally a secure method for sharing clinical research data
- You do not need to consider using a sharable data dictionary and standards if you plan to do the analysis yourself.
- Before sending your data to collaborators you should check that it contains no personal identifiers.
Q7. Double data entry is a good method for data quality control and it works as follows:
- A member of the study team enters the same data into two separate patient records.
- Two members of the study team enter the same data for the same patient record which is then compared for concordance.
- A member of the study team enters the data into two separate electronic systems.
- Two members of the study team alternate data entry for a single patient throughout the study.
Q8. Which of the following would you consider to be the most appropriate way to collect data on drug use that is entered remotely by the study participant?
- Giving the participant a set of blank paper forms (no envelopes) that they have to drop in your mailbox or hand over to someone traveling in your direction.
- Contacting them from an unknown phone number, but using an interactive voice response (IVR) system.
- Sending the participant an SMS message from a number that is unknown to them and prompting them to text you back their information.
- Sending the participant a link to a web form on a secure website belonging to the research study group.
Week 02: Data Management for Clinical Research
Q1. Standard Operating Procedures (SOPs) are “detailed, written instructions to achieve uniformity of the performance of a specific function.” Which of the following statements is FALSE regarding SOPs?
- Authoring SOPs is enough to ensure that your study is in regulatory or contractual compliance.
- The development of SOPs while planning your study helps you define the operation and technical components of your project.
- Authoring SOPs can help you define the milestones on your study timeline and reveal potential roadblocks.
- Operational metrics and explicit SOPs can help your study be transparent for external review and provide a framework for accountability within your study team.
Q2. Which of the following guidelines for creating SOPs were presented in the lecture video? (Check all that apply)
- SOPs are often needed for explaining and justifying unique processes to external groups that will approve, fund or audit your study.
- SOPs define best practices and as such whether the the study personnel sign off on them is irrelevant.
- Committing sufficient time to SOP development is a good idea and will help your team derive more value from your SOPs.
- SOPs should be self-explanatory, standalone documents.
Q3. Read the following statements about using validated instruments for your data collection. Which statement is FALSE?
- Even if an instrument is validated in an open peer reviewed journal you will still need to verify that there are no restrictions (licensing, attribution, etc.) prohibiting you from using it.
- An instrument is considered reliable if the results you derive from using it are reproducible and consistent.
- An instrument is considered valid if its ability to measure what it claims to measures has been tested via a dedicated study.
- If the validity of the an instrument has been established in a peer reviewed publication, you do not need to review the validation methods and context before utilizing it your study.
Q4. Which of the following hypothetical terminologies that would be used for an emergency department’s registry of bone fractures has the highest granularity?
- A list of a lay terms describing the body part where a fracture occurred (neck, head, left arm, right arm, left thigh, etc.)
- A list of all the anatomical names of human bones to indicate the specific bone where a fracture occurred.
- A terminology composed of two parts: the first term would be picked from a list of all the anatomical names of human bones and the second term would be picked from a controlled list of terms that describe the types of bone fractures: simple fracture, compound fracture, etc.
Q5. The 13 core principles of Good Clinical Practice (GCP) map to what two overarching concepts? (check all that apply)
- Respect for the rights of study participants
- Cost efficiency of research planning
- Conduct of accurate and verifiable research
- Publishing results in a high-impact journal
Q6. In what scenario do you need to comply with 21 CFR Part 11 when planning a study for FDA submission?
- When study personnel sign paper forms generated by an electronic system.
- When the study source documents are entirely electronic, with no paper.
- When you use any computer system in a clinical study.
- When study documents are scanned into a computer system.
Q7. Electronic Data Capture systems can log all data viewing and manipulation events to create an audit trail. Why are audit trails important? (Check all that apply)
- Audit trails are really only useful in limited clinical research scenarios like randomized controlled drug trials.
- Logging access and the history of every data point collected and changed throughout the lifetime of the study helps with transparency and accountability in your study.
- Audit trails are no longer required or even needed for research studies.
- An audit trail is often necessary to fulfill regulatory compliance requirements set by external agencies.
Q8. Electronic Data Capture systems allow study personnel to enter research data into the database via computerized Case Report Forms (CRFs). Which of the following functions is NOT desired in a e-CRF?
- Displaying the variable names in the database that correspond to the questions on the CRF.
- Providing branching logic that alters the flow of the form based on responses to specific questions.
- Forcing the users to chose only from the given options for categorical variables.
- Using data validation that prevents the users from entering values in given fields unless they conform to the data dictionary. (e.g., preventing malformed dates for date fields, checking numeric ranges, etc.)
Q9. An Electronic Data Capture system will need to allow users to export the study data. Which of the following statements do you agree with based on the content of the lectures? (Check all that apply)
- The data exported from the system will most likely need to be analyzed. Therefore the systems should have capacity to export the data in formats that can be consumed by most data analysis software.
- Date-shifting is one technique for de-identifying date values upon export.
- An EDC system should be able to email the raw data directly to anyone on the internet that the study owner chooses.
- The EDC system can help preserve subject privacy by giving the end-users a means to de-identify data upon export.
Q10. Electronic Data Capture systems can empower study investigators to create and manage their own projects without relying on software programmers. Which of the following features reflect that? (Check all that apply)
- Study team members can create the underlying database and the CRF elements using an intuitive (point and click) interface.
- The project owner can add and remove specific users and assign them permissions that match their role in the study
- Members of the study team can create the underlying database and CRF elements using a metadata file that they can re-use or share with collaborators.
- Members of the study team can set up Electronic Data Capture projects, but only if they document the purpose of their data capture study using a lengthy and detailed questionnaire.
Week 03: Data Management for Clinical Research
Q1. Before answering these quiz questions, please download and read the hypothetical clinical research scenario (PDF linked below), which includes a partial data collection plan.
- Got it!
Q2. Based on what you have learned so far, what would be the best approach to capture the age of the children for purposes of this study?
- Ask for the age in days at baseline and in each of the case report forms during febrile illness episodes.
- Ask for the age in months at baseline.
- Ask for the age of the children in years in the baseline form. It will be more or less the same throughout the one-year duration of this study.
- Ask for the date of birth in the baseline form. Make sure that the baseline form contains a date field which documents the baseline date for this one-year study. The febrile illness CRFs will also contain the dates of those corresponding illnesses.
Q3. Based on the study design and the young age of the children the best place to ask for height is in the:
- Baseline form only.
- Baseline form + febrile illness episodes CRF.
- Febrile illness episodes CRFs (but not in the baseline form).
Q4. Which of the following would be essential to how you capture the weight of the children throughout the different events of this study? (Check all that apply)
- A text field that is validated as a numeric field type.
- A human readable label in the appropriate form that lists the unit (e.g. kg) of the requested weight measurement.
- The associated date field within the form (e.g. baseline date or case report date) which can then be associated with each particular weight measurement during analysis.
Q5. From the following choices, what would be the most consistent way to measure the duration of illness if possible?
- The number of days that the child did not attend daycare during that illness episode.
- The number of days in which the child’s temperature was measured at over 38 degrees Celsius.
- The number of days that the parents report the child had fever.
Q6. The investigators who designed this study want to study the temporal relationship between febrile illness episodes and the children’s flu vaccination. Specifically they are interested in the interval (in days) between the beginning of the illness and the vaccination date. What is the best way to achieve that?
- You do not need to capture this interval explicitly in the forms. You can compute this information from already captured dates: (1) the date of the vaccine in the baseline form and (2) the date associated with each febrile episode CRF.
- You can record in a numerical field in every illness CRF the number of days that have passed since the child received the flu vaccine.
- When a child is vaccinated, you enter the date of vaccination in the baseline form. Then you go back to every febrile illness CRF that preceded the vaccination and enter the interval (in days) between the start of those illness episodes and the vaccination date.
Q7. Assume that experts in this field have determined a comprehensive list of associated symptoms that need to be captured in the case report forms. Assume that you also need to include a free text option to include additional symptoms that your study personnel may chose to capture. What is the best way to capture the child’s associated symptoms in your CRF?
- Drop down field with all symptoms as options in addition to the option “other”. Set up branching logic that shows you an additional free text field if the user selects “other.”
- Radio button field with all symptoms as options in addition to the option “other”. Set up branching logic that shows you an additional free text field if the user selects “other.”
- A check box that allows you to check all the symptoms that apply as well as the option “other.” Set up branching logic that shows you an additional free text field if the user selects “other.”
- Free text block where the study personnel can enter all the symptoms in an empty text area where they write each symptom on a new line.
Week 04: Data Management for Clinical Research
Q1. We should aim to avoid mid-study data changes, but sometimes
they are necessary. What are some good reasons for mid-study data changes?
(check all that apply)
- A scientific discovery adds to the types of data you need to collect.
- Your team would like to test an exciting new piece of software.
- A natural disaster destroys your lab and disrupts your study schedule.
- Your lead data manager goes on holiday.
Q2. What should you AVOID when making mid-study data changes? (check all that apply)
- Recoding data.
- Deleting data that you’ve collected that you’re not going to use.
- Changing the meaning of data fields.
- Collecting newly inserted data fields retrospectively when possible.
- Versioning data collection instruments
Q3. What types of study events require immediate IRB review (in US settings)? (check all that apply)
- Mid-study data analyses
- Data privacy breaches
- Protocol modifications
- New patient enrollment
- Changes in key study personnel
- Serious adverse events
Q4. What study activities are often reviewed by a Data Safety Monitoring Board? (check all that apply)
- Secure network configuration of electronic data collection systems
- Quality and completeness of the study data
- Statistical reports based on interim datasets
- Hiring of study personnel
- Participant enrollment and retention at different study sites
Q5. You are working on a retrospective study analyzing emergency department visits between 2005 and 2012. However, the dataset you have received only includes data up to 2010. In the context of your study, this dataset rates poorly on what dimension of data quality?
Q6 .In a longitudinal clinical trial, how do missed lab measurements affect your study data completeness?
- Missed lab measurements have no effect on overall data completeness.
- It is impossible to miss taking lab measurements, because these activities are specified in the study protocol.
- Missed lab measurements permanently affect your data completeness, as they were not captured during the measurement window and never will be.
- Missed lab measurements can temporarily reduce the completeness of your study data, but lab measurements taken later can be used to fill in the gaps.
Q7. A data audit process that compares paper source documents to the study database will probably not detect which problem?
- Personnel are not entering data promptly into the electronic system.
- The transcribed data are full of typographical errors.
- Source documents are being discarded after data entry.
- Lab technicians are not calibrating lab equipment routinely.
Week 05: Data Management for Clinical Research
Q1. What types of research publications must be deposited in PubMed Central?
- Only papers that are published in non-Open Access journals.
- All papers resulting from work funded by the NIH.
- All papers on human subjects research.
- Reports of clinical trials only.
Q2. Which data elements listed below are examples of study Identifiers according to HIPAA? (check all that apply)
- Car license plate number
- Weight at study baseline
- Study enrollment date
- Mother’s name
- Country of residence
Q3. Why is date-shifting a popular method for de-identifying dates in study datasets? (check all that apply)
- Date-shifting preserves the order of recorded events.
- The process automatically de-identifies dates that identify people 90 years of age and older.
- Date-shifting doesn’t require creating a new set of variables and updating your codebook.
- Date-shifted data clearly looks like de-identified data.
Q4. You are conducting a study that evaluates patient adherence to prescribed medications. Which of the following electronic systems could you use as possible sources of data? (Check all that apply)
- The lab system for monitoring blood drug level.
- Physician and nursing documentation systems from subsequent visits.
- Order entry and/or prescription writing system (Rx Writer).
- Pharmacy system if some or all medications are dispensed by your institution’s pharmacy.
Q5. Which of the following are study data-related problems that stem from using aggregation systems? (Check all that apply)
- You may not be getting real-time data.
- You may not be getting complete data.
- The data are never integrated across multiple systems.
- If a production system is heavily used in the clinic (such as a scheduling or a billing system), your query to the aggregating system may affect or be affected by real-time usage volume of the production system.
Q6. Which of the following statements are true about Picture Archive and Communication Systems (PACS)? (check all that apply)
- PACS are used for transferring image files between scanners. Images are eventually archived in a high-capacity scanner.
- PACS provides a read-only repository for storage and access of acquired imaging data.
- PACS typically store images using the DICOM (Digital Imaging and Communications in Medicine) file format.
- Due to its large memory capacity, a PACS is the ideal machine to run complex image analysis algorithms.
Q7. The advantages of using a de-centralized data hosting model for multi-center studies are (check all that apply):
- The data collection sites retain the ability to store and manage the data locally.
- Data can be collected using whatever local data model is in place at each site. This eliminates the need for study personnel to extract, transform, and load (ETL) the data into a common standard for analysis.
- Unlike in the centralized data hosting model, sites cannot see data from other sites.
Q8. Which of the following considerations for form design require understanding of your local settings? (check all that applyt
- High illiteracy rates means that many participants cannot spell their name which could lead to different phonetic spellings by different study personnel.
- Incomplete record keeping around birth may mean that participants cannot accurately or consistently report their exact day of birth.
- The address or place of residence of transitory populations may require that you structure your data collection to capture that information more frequently.
- You should always use paper forms and not computers or tablets to avoid the appearance of wealth.
Week 06: Data Management for Clinical Research
Q1. Which of the following are advantages of using surveys to collect data directly from research participants? (Check all that apply)
- Surveys help you avoid interviewer bias especially around sensitive topics.
- In the United States, you do not need IRB approval to conduct anonymous clinical research surveys.
- If survey instructions are not clear, it is not a big problem because survey participants can ask for clarification.
- Participant responses can be completely anonymous.
- Direct data entry (especially if the survey instrument is electronic) eliminates data entry errors by study personnel.
Q2. Which of the following recommendations about survey questions were made in the video lectures? (Check all that apply)
- Sometimes you and or your colleagues can think of interesting things to ask your respondents that are outside the scope of your specific aims. Adding those questions to your survey will give you additional information without any negative impact.
- You can condense questions and minimize respondent fatigue by creating compound questions that ask about multiple concepts which can be combined with “AND” or “OR”.
- Avoid acronyms and technical jargon unless it is appropriate for your audience.
- Use simple grammar and neutral, non-biased terms.
Q3. Recall bias … (check all that apply)
- Is due to people’s tendency to lie about events that happened a long time ago.
- Can be minimized when the recall questions address a shorter and more recent time period.
- Is a poorly understood phenomenon.
- Can be minimized for the recent past using short series of related questions.
Q4. Which of the following statements about survey questions is true based on the video lectures. (Check all that apply.)
- Having branching logic in your survey increases survey fatigue.
- Visual Analog Scales (VAS) can be represented as slider questions when administered online. VAS can minimize respondent perception and memory bias.
- Likert Scale questions should be balanced, should have equal distance in the gaps between the choices, and should have an even number of choices.
- Branching logic is one way to increase the validity of your data.
Q5. When you assemble your data collection instrument, which of the following actions is likely to annoy, overwhelm, or discourage your participants?
- Don’t reveal the real amount of time that it will take to fill the survey. Always under-represent the real amount of time to trick your respondents into starting the survey.
- Mentioning up-front all the sources of information they will need to access to be able to answer your survey.
- Include ample information at the beginning of the survey explaining to survey respondents the intent and purpose of the survey.
- Giving detailed explanation of the steps you are taking to ensure their anonymity in anonymous surveys.
Q6. It is important to perform pilot testing of your surveys before you begin the study. The pilot test should be conducted in a group of participants outside your study team who represent your target population. From the data perspective, this pilot testing is important because… (check all that apply)
- You may find some invalid or inconsistent data in your pilot responses and can anticipate and correct that in your instrument.
- If you’re lucky, you may end up using this sample population’s data to draw conclusions about your study hypothesis thereby saving the cost of the larger study.
- The pilot testing may inform you whether or not there will be variance in question responses (e.g. in Likert Scales)
- You may find out that many of the respondents do not have information at hand to answer some of your questions and you will therefore be informed about how complete your survey responses may be.
More Google Cloud Free Courses
More Free AWS Courses