Pharmacy Student Scholarly Project: Data Management

Resources to help you complete your required Pharm D Scholarly Project.

The Norris Medical Library cannot provide instruction or help with data collection or data analysis, including selecting and using software; ethical/legal issues with data collection; creating surveys; inputting data into software; selecting and running statistical analyses and tests; or determining significance of data. The resources below are provided as an introduction to your self-education in these concepts.

What is data management?

Data Management is a process that encompasses:

  • deciding what data to collect for a research project
  • deciding how to capture the data, or finding already-collected data that you can use
  • selecting a tool for data collection and setting it up
  • selecting a software for data analysis and learning to use it
  • following applicable laws and guidelines when collecting, storing, and analyzing data
  • creating mechanisms to allow others to access and use your data. This step is optional and not all projects undertake it.

Data management should be a part of your project planning from the beginning.

Data Collection

Your research project might include collecting data from individuals or groups. Using software to collect this data ensures uniformity; software also allows for export of data into data analysis software. You may also choose to use paper forms for data collection and later enter the data into software. To choose whether to use an inexpensive software (e.g. SurveyMonkey; Wufoo; Polls for Pages Facebook app, etc.), a USC-provided subscribed software (e.g. Qualtrics, RedCap), or paper forms, consider:

  • The types of data you are gathering and laws or ethical codes regarding protection of that data. Review the USC Office for the Protection of Research Subjects' Internet Research page and Responsible Conduct of Research page for information on laws and ethics.
  • How the survey or form will be completed, and who will complete it. If you will be filling out the form while walking around a pharmacy, it may be useful to have a form you can fill in on your phone. If you will have all study participants fill in the form themselves at their homes, you would need it to be compatible with a wide variety of computers/smartphones, or you may decide to use paper forms and enter the data into the software yourself.
  • Data collection software export features and compatibility of the exported file type with your data analysis software

Creating surveys

"Writing and administering a survey" is a frequent research method. However, writing a survey is deceptively complex. How do you know that your questions and answers accurately capture people's opinions? How do you know that the answer choices you have provided are readable and actually able to be analyzed? 

  • A Validated Survey is a survey that has been tested and determined to actually and accurately capture the data it aims to get-- a validated survey is a reliable research tool. Many researchers who create validated surveys release these surveys for future use (sometimes this is free, sometimes you must pay a fee). Re-using a validated survey can also link your research with existing research, increasing its value. Use the databases below to find surveys that others have created for re-use. Note that not every survey is validated-- read the article describing the development of the survey to find out.
  • Review the books below for tips on survey writing, including getting data that can be analyzed. It is time-consuming and expensive to validate a survey, but you can test it on friends and family members to ensure readability and understanding. You can also use the data from your tests to determine if it is analyzable.
  • To consider whether a question in "analyzable," consider this question and answer:
    Question: How many texts do you send per day? Answer choices: A. 0-5 B. 5-10  C. 11-30  D. 31-35
    If a person texts 5 times a day, they could pick choice A or choice B. It will be hard to analyze your survey if 5-times-a-day texters can be in two separate categories.
    The answer choices also are not balanced. You will likely have more respondents in category C because that includes people who text 11 through 30 times per day; every other category includes only 5 numbers (0-5; 5-10, 31-35). This type of unbalanced response set can mislead you when analyzing data.
    This question is analyzable; the answer choices could be re-written to make this more easily analyzable.

Finding Free-To-Reuse Data

You may not need to collect data yourself. Using already-collected data can help you move quickly into data analysis. The sites below include freely-available data sets that you can download into data analysis software. In addition to these sites, think about government or non-governmental agencies that may collect data on your specific topic, and use Google to find their websites and look for data.

Data Analysis

Use data analysis software to analyze your data.

All PC computers at the Norris Medical Library have Microsoft Excel, SAS 9.4, SPSS 24, and STATA 14 installed. The library does not offer training or support on using these programs. Use the links and books below to learn about more ways to access these programs and learn to use them.
 

The Norris Library also owns hundreds of books on data analysis. Search the library catalog for the words "data analysis," the name of the data analysis program you are using, or the name of the type of test you want to learn about, to find books on these topics.