Open Science Mini-session

class: center, middle, inverse, title-slide

# Open Science Mini-session
### Thomas Pollet (<span class="citation">@tvpollet</span>), Northumbria University
### 2018-12-05 | <a href="http://tvpollet.github.io/disclaimer">disclaimer</a>

---

## Open Science

* Following troubling reports, especially in social psychology.

* Umbrella term.

* Reform -- Center for Open Science.

* Platform.

---
## My experience.

Started using it for MSc. projects. Especially for replication projects. One of them published
<a href="https://www.sciencedirect.com/science/article/pii/S0141938218300349" target="_blank">here</a>.

Response is overwhelmingly positive.

Students engage with it.

Goal setting.

Product for student to market themselves with.

---
## Why bother?

* More and more reviewers ask for this. (Journals now have registered reports)

* More and more job adverts ask for this.

* Good practice.

* OSF framework is tool for project management and version control.

* Marketing tool for showcasing students' work!

* [Preprints](https://psyarxiv.com/).

---
## A quick tour.

**Thomas shows the <a href="https://osf.io" target="_blank">OSF</a>.  **

---
## Preregistration.

* Open-ended registration. - Just write what you want!

* Registered report.

* aspredicted.org (9 questions)

* OSF standard (3 questions!)

* Some specific standard forms for _replication_ studies.

--> the emphasis is on _pre_-registration. Before you collect data, you should make key decisions.

--> Not that much work, per se, as key information is also what you should have in your ethics application.

---
## Aspredicted.org

Let's look at a completed example from the [aspredicted website](http://web.archive.org/web/20160407104112/https://aspredicted.org/nfj4s.pdf).

---
## Things to consider.

Some forms will ask you this upfront:

* How many conditions (_a priori_)?

* When will you stop data collection?

--> What if you change your mind?

You can put a note why you diverged or set up a new preregistration.

---
## What about qualitative data?

* Feasible... ?! The basic premise of setting out what you are going to do, should not be hard.

* Can be relatively open-ended. To a large degree your approved Ethics protocol already has most info.?

* Likely in the future open science practices. Check [this](http://blog.ukdataservice.ac.uk/show-me-the-data/)!

* Template available [here](https://osf.io/j7ghv/wiki/home/).

---
## Criticism: Stifling of creativity?

Not necessarily: just differentiation between confirmatory and exploratory!

---
##Further reading.

* [Making replications mainstream](https://www.cambridge.org/core/journals/behavioral-and-brain-sciences/article/making-replication-mainstream/2E3D8805BF34927A76B963C7BBE36AC7)

* Annotated reading list: [https://psyarxiv.com/cfzyx](https://psyarxiv.com/cfzyx)

---
##Any Questions?

[http://tvpollet.github.io](http://tvpollet.github.io)

Twitter: @tvpollet