PH30004 Data Gathering/analysis Plan
Purpose and description
The purpose of this assignment is to provide you with an opportunity to identify and describe a plan to gather and analyze data to address a research question of interest. You are also asked to consider potential challenges that might impact this process.
Course learning outcomes
5. Communicate the relationships among a study research question, research design, data, measurement, and data analysis (Concepts 3 and 10)
Assignment Requirements
This assignment may be completed on a group (2, 3, or 4 people) or individual basis. If turned in as a group assignment – one group member must upload to Canvas, and all group members names must be on the uploaded document. You can opt to do this assignment in a group even if you did not complete the prior research design assignment as a group.
- Provide a public health-related research question of interest. You may use a class-identified question or a question of your preference. You may use the same question you used for the previous research design assignment.
- Identify a design choice. If you are reusing your question from the research design assignment, you may use the same design, or you may use something different. You do not need to describe or provide a rationale for this choice, just name the design.
- Identify the type of data you would need to gather to address this question, that is consistent with the design. Describe this with as much detail as possible. For example, if you want to conduct group interviews, your data will probably be written notes or audio or video recordings of participant responses. Describe not just the data (responses) but what format it will be in (audio recordings, notes, typed transcripts, etc.)
- 4. Describe how you will get this data, in particular what sample or population of participants you will need. You do not need to describe an identification or recruitment process. You should describe inclusion criteria – what data will work. You might also describe exclusion criteria – what data or participants will not work You do not need to describe all inclusion and all exclusion criteria, but you should include the most important items. For example, if your group interviews are about environmental concerns in Cleveland, your sample will include people who know about or are impacted by this issue – so this is a criterion for inclusion. Depending on your specific question, you may want to recruit residents, or you may want to recruit public officials and policy makers. You may also want to include others such as scientists or business leaders. But data from people in Akron, or Kent, even if convenient, is not necessarily going to help you understand concerns in Cleveland. So “Living in Cleveland” or “living or working in Cleveland” are possibly inclusion criteria. As another example, if you want to use secondary data analysis to identify trends in pregnancy among adolescent girls, then you need a dataset that includes information about pregnancy in women aged 13 to 17. You could also use data from women who are 18 or older, as long as the dataset includes historical information, i.e., data that show whether they were ever pregnant when aged 13 to 18. But data about men, or young children, or data about adults that is not historical/retrospective, is no use. These things – age, sex, pregnancy status – describe inclusion or exclusion criteria. You can combine the information in items 3 and 4 however you like. You may describe the sample/participants first (item 4), then the data, if you prefer.
- 5. In very general terms, describe one way to analyze the data you get. For example, in an experimental or trial design, you might want to compare results of a treatment and control or delayed treatment group. Most likely you will use statistical analysis of measures of treatment outcome. If the treatment is meant to lower blood pressure, you might use a two sample t test to compare blood pressure between the treatment and control group. Alternately, for survey-type questions such as “how many people prefer A over B?” – then simple proportions, otherwise known as descriptive statistics, are all that are needed, to compare the proportion of respondents who picked A to those who picked B. This analysis consists of just counting responses and figuring out what percent picked each choice. On the other hand, if you are using individual or group interviews, then you might use basic coding (you do not have to provide detail about types of codes or meaning units) or you might instead use frequency-based content analysis, such as a word cloud.
- Lastly, provide in a sentence or two, what you think is the most challenging part of this data/analysis process. Is it recruiting participants? Finding existing data? Learning how to run statistical models? Time it will take to do data analysis? Do you have concerns about the quality of data, for instance, if you propose an experiment, are people likely to drop out before the treatment is over? You do not need to provide a solution, just identify the challenge.
Please address these items on a single document. The document should be no more than 1.5 pages of single spaced or 3 pages of double-spaced text in length. You may write in paragraph form or present your responses to each item, using the numbering shown above. You do not need to do research on data analysis methods – describe things based on your current level of knowledge. You can also look ahead in the Social Science Research text to chapters 13-15 that are about data analysis, that are assigned reading next month, and use some of the information provided. You can use external sources but are not required to do so. Please cite any sources you use. If you directly quote from a source, please be certain to use quotation marks and show the specific citation information.