Week 8; Open Science Flashcards
What is Open Science?
Transparent and accessible knowledge that is shared and developed through collaborative networks
Open Access
-Free access to the papers that describe the outcomes of research
- No fees to read, cite or criticise research
Open Data
-Sharing data so others can perform important practices such as meta- analyses
- Also enables other researchers to reproduce your findings
Open Practices
-Providing detail about the procedures, techniques and statistical analyses involved in the research and thus allowing for the greatest transparency in research communication
- Specifying the protocols before analyses of the data commences aids transparency
Open Collaboration
- Supporting researchers to take part in prepublication sharing
Data Dredging/ P Hacking
The misuse of data analysis to find patterns that can be presented as statistically significant when in fact there is no real effect
How to Torture Data
- Using creative- outlier rejection
- Selective Reporting
- Post-hoc Theorising
-Analysis of sub-groups - Not adjusting for multiple comparisons
Outlier Rejection
Identifying and removing data points that significantly differ from the rest of the dataset
Selective Reporting
Certain results from scientific research are deliberately not fully or accurately reported, often to suppress negative or undesirable findings
Post Hoc Theorising
Generating hypotheses based on data already observed
How to avoid torturing data
- One tailed hypothesis tests are discouraged
- Rely on effect sizes instead of p-values
- If a variable is continuous, keep it continuous
- Educate yourself on how to handle missing data
- Use covariates
Benefits of Open Science
- Makes science available to as many people as possible
- Assists scientists to say transparent and honourable
P-Hacking
Analysing your data in multiple ways and selectively only reporting those that result in a p-value that is smaller than 0.05
Financial Conflicts of Interest
The researcher is receiving funding from a relevant company i.e. pharmaceuticals may fund research on their competitors
Professional Conflicts of Interest
A researcher may only receive a promotion if their findings are statistically significant
Personal Conflicts of Interest
-Someone that the researcher knows may be personally involved in the topic of research
- The researcher may hold a strong opinion that the research would/ would not report
