“Open as standard” policy at UiO
- “Scientists and students are responsible for managing research data according to the principles and requirements stated above. Supervisors of ph.d candidates and students have a special responsibility for ensuring that candidates and students attend courses and manage research data according to the above guidelines.”
Check out the following open science tools by clicking on the '+';
Preregistration
What is it?
- Preregistration means to register a time-stamped copy of your a priori hypotheses prior to conducting the research. Preregistration can be embargoed while the study is being conducted.
Why is it helpful?
- Preregistration is not only required by more and more journals, but most of all it helps you in planning your research more effectively and rigorously. By thinking about your hypotheses prior to conducting the research and registering these, you usually make a more thorough literature search prior to setting up the study, you are more aware of prior research, decide on the theoretical framework that you aim to use, and as a result can better position your planned research and its contributions. Doing so, will, for instance, make you aware of relevant control variables or additional variables to rule out alternative explanations. Likewise, thinking about the required sample size and type of analysis as part of a preregistration helps you in properly planning your research and increasing the likelihood of finding effects – if they exist. Additionally, even if your hypotheses are not supported, publishing these is easier compared with attempting to publish null-findings from not preregistered studies. Preregistration also helps reviewers and readers evaluate the veracity of your statistical analysis. You can also preregister research questions and additional variables for which you do not have a priori assumptions and state that you will test these in addition to your hypotheses. For more on the benefits of preregistration in psychology, see this paper: Lakens (2019).
How can you do it?
- Several websites to preregister your studies exist. Try out aspredicted.org, which asks you all essential questions that you should think about as part of its preregistration flow, or check out the preregistration option on the Open Science Framework (osf). You find more information on how to preregister on osf here.
- You are not sure how to preregister special forms of studies? The following links might be helpful:
- Template for preregistering a correlational meta-analysis
- Template for preregistering an experimental meta-analysis
- Examples:
- Preregistration of analyses of existing data: Steffan et al (2021), Rosslund et al (2023)
- Preregistration prior to data collection: Kartushina & Mayor (2020)
A Priori Power Analysis
What is it?
- A priori power analysis calculates the number of participants needed to be sampled given an expected effect size and study design.
Why is it helpful?
- By calculating the required sample size to find the expected effect, a priori power analysis maximizes the likelihood of finding an effect that exists in the population. Without a power analysis, you can otherwise never be sure whether not finding an effect means that the effect is not there or whether it is there, but you could not detect it.
How can you do it?
- There are numerous ways of calculating the required sample size given your study design, expected effect size, and desired power. To estimate the expected effect size, you can use the effect sizes of previously published studies, one of your pilot studies (if you conducted one), or make a guess whether you would expect a small, medium, or large effect. Simulation studies (for instance, in R) can further help in calculating the required sample size. For more on how to estimate your expected effect size of interest, see Lakens (2022).
- There are a number of tools that can help you in performing your power analysis. Power analyses for the most common study designs can be done in the cost-free tool G*Power by Faul et al. (2007), which you can download together with extensive documentation here. There also exist a couple of shiny apps for specific study designs not covered in G*Power, such as this one developed by Daniel Lakens.
- How to perform power analyses using simulations: Power analysis workshop by Christopher M. M. Cox
Registered Reports
What is it?
- Registered Reports replaces the submission of complete manuscripts for publications. Instead, they split the submission into two phases. First, and to conducting the research, a manuscript including research question, theory, methods, analysis plan including pre-registration is submitted. The journal organizes a peer-review process and, possibly after revisions, gives an in-principle-acceptance (IPA). If the researchers conduct the study as detailed, the results will be published irrespective of the outcome in the final manuscript. The registered report format can be used for both replication and original studies. Deviations from the registered methods and analysis plan are possible but need to be fully transparent, see Lakens (2024).
Why is it helpful?
- For the researcher, Registered Reports are an excellent way to benefit from preregistration. It frees them from the pressure to find hypothesis-confirming results. For the community, it decreases publication bias in the form of suppression of negative results. For comparisons of results of registered reports and traditional reports, see Scheel et al. (2021).
How can you do it?
- There are various possible publication outlets. Some journals accept registered reports along with other manuscripts (e.g., Nature Human Behavior). Other journals accept only registered reports (e.g., Comprehensive Results in Social Psychology). Finally, Peer Community In Registered Reports (PCIRR) organizes a process of reviewing and recommending registered reports that are then accepted by PCIRR-friendly journals without further peer review.
- The Center of Open Science keeps a list of journals that accept registered reports; another one is here. But make sure to also check whether your preferred outlets accept them.
- There is funding available specifically to support registered reports, e.g. Stiftelsen Dam
- Many researchers at PSI have published registered reports previously, for instance:
- Askelund et al. (2024) in BMC Medicine, a Registered Report based on secondary data
- Carlyle et al. (2023) in Addiction Research & Theory using PCIRR
- Kartushina & Mayor (2022) in Developmental Science
- Zickfeld et al. (2018) in Comprehensive Results in Social Psychology showing how one registration can span multiple studies
Open Lab Notebook
What is it?
- Open Lab Notebooks are a way to post and share experimental data and research protocols in real-time.
Why is it helpful?
- This approach can provide transparency for the scientific process
How can you do it?
- The use of Jupyter is one approach for maintaining an open lab notebook. See Sprengholz (2018), for an example implementation.
Open Analysis Code
What is it?
- Open Analysis Code refers to making the analysis code that is used for preparing and analyzing the data set publicly available.
Why is it helpful?
- It can make it easier for others to better understand and reproduce your analysis
- Open analysis code might be cited by others in future research
- Open analysis code also makes it easier for you to reproduce your own analysis in the future
How can you do it?
- How to create reproducible research reports using RMarkdown: Reproducible research reports
Open Materials
What is it?
- Open Materials means that all study materials (e.g., questionnaires, experimental manipulations) are made publicly available.
Why is it helpful?
- Open materials can help others better understand your method and can be re-used in future research
How can you do it?
- Open materials can be posted on open science framework
Open Data
What is it?
- Open Data refers to openly sharing the (anonymized) data used in the analysis as long as participants consent to sharing their anonymized data. Thereby, a general recommendation is to share as openly as possible but to be as closed as necessary.
Why is it helpful?
- Open data can facilitate the verification of results reported in papers
- New hypotheses can be generated using the novel analysis
- Integration into future meta-analyses or individual patient data (IPD) meta-analyses
How can you do it?
- The importance of open data for developmental science: Open Data for Developmental Science
Open Acess
What is it?
- Open Access is a publishing method, whereby the manuscript is made available and hence accessible to all researchers and the public for free and without being hidden behind a paywall.
Why is it helpful?
- Increases the accessibility of your work
- Open access articles receive more citations, on average
How can you do it?
- See the UiO library open access page for resources on how to publish open access
Replication Study
What is it?
- A Replication Study is a study that tests whether the results can be reproduced in another study that follows the same study setup. Replication studies can be generally categorised into two types: direct replications (the methods are are close to the target study as possible) or conceptual replications (there are some differences in the methods compared to the target study)
Why is it helpful?
- This can provide more confidence in a finding and can be useful for evaluating the generalisability of results
How can you do it?