Dr Hannah Fraser
@HannahSFraser
Followers
1K
Following
9K
Media
96
Statuses
3K
Research working for UQ on birds and @unimelb on meta-research on the @repliCATS project. Views expressed are my own. Retweets =/= endorsement She/her
Joined January 2014
RT @SABE_latrobe: šOur daily decisions are inextricably linked with the planet's health. While governments at @COP16Oficial are working toā¦.
0
6
0
In my opinion, we need to move towards a norm in science where we analyse our data in a way that tests the robustness of our findings and present and interpret evidence with humility and nuance.
0
0
2
If asked I'm sure 'we' (scientists) would agree that no single study or result is definitive but the incentives for conducting novel research contribute to a culture where published research is considered final and repetitions are considered uninteresting and unpublishable.
1
0
2
So honoured to have this story coming out about our work. If results vary so much according to who analysed the data, we need to be a bit more careful about how we conduct and interpret research.
>200 scientists analyzed the same data, the results varied widely. it shows āwe canāt be relying on any individual result or any individual study to tell us the whole storyā @HannahSFraser says. w/ @TimParker88 @Elliot_Gould_ @Nicole_C_Nelson @BrianNosek .
2
6
34
In incredible news, our preprint ( has now been downloaded >7500 times from EcoEvoRXiv in the last week! We show that ecological results are very sensitive to who analyses the data and how they do it.
So excited to release this work into the wild. Thank you so much to the 300 odd researchers who contributed to the project. And especially to my joint-first-authors @TimParker88, @Elliot_Gould_ for putting in a terrifying number of hours on this.
1
16
55
@TimParker88 @Elliot_Gould_ We also checked whether the uniqueness of the combination of variables included in the analyses (mean Sorensen's index) influenced the likelihood that a particular analysis found a very different result. there was no convincing evidence for this either.
0
0
7
@TimParker88 @Elliot_Gould_ Another subset of our participating researchers were in charge of reviewing the methods of all the analyses and scoring them. We expected that there might be a relationship between having a low review score and finding a really different result. There. wasn't really
3
1
8
@TimParker88 @Elliot_Gould_ We gave researchers 2 research questions "how does grass cover influence Eucalyptus seedling recruitment" and "to what extent does sibling competition influence blue tit growth" with matching datasets and asked them to analyse them and send us their results and methods.
2
3
11
So excited to release this work into the wild. Thank you so much to the 300 odd researchers who contributed to the project. And especially to my joint-first-authors @TimParker88, @Elliot_Gould_ for putting in a terrifying number of hours on this.
When many different researchers analyze the same data, results vary . A LOT!. Our Many-Analysts study from ecology and evo-bio is FINALLY up on @ecoevorxiv . Thanks to the HUNDREDS of researchers who made this study possible.
2
16
44
Phd student @danielghamilton is seeking Aussies whoāve been diagnosed with cancer to do a quick survey to help us understand how they feel about researchers sharing (or not sharing) anonymous cancer research data with other researchers.
If you participated in cancer research, how would you feel about your anonymous data being shared with other researchers? What about the public?. We want to hear from Australians previously diagnosed with cancer. Tap to learn more.
0
0
2
RT @KatheMathBio: Interested in soil data? Iām pulling together a series of 3 workshops over the summer on #SoilData (data models, templateā¦.
docs.google.com
Bridging soil data, knowledge, and semantics Registration - https://forms.gle/MqJfTpBQXQeij8Fr6 DONE and Archived see session ārecapā for report outs Point of contact: Kathe Todd-Brown (University of...
0
24
0
RT @betsylevyp: š¢Sending out an SOS to all pollsters / social scientists: Want to know whether the SCOTUS decision will change abortion opā¦.
0
248
0
RT @redpenblackpen: What kinda of funny things might you do returning to the office after two years? Like desk excavation or cleaning the oā¦.
0
1
0
RT @moorterb: 3 year postdoc vacancy on species distribution modelling and connectivity in Norway with @BobOHara and @joechip90 for detailsā¦.
0
54
0
RT @StacyTShaw: Today has been terrible, and because I've already seen so many adorable pictures of your pets, please tell me about the cooā¦.
0
1
0
Aptly named 'F yoU pay'. Really brings us back to our roots where people were sent on. A dangerous ship journey to a new colony for difficulty affording these same items. This is not a good thing.
Fupay will become the nationās first buy now, pay later provider to offer short-term loans for items such as milk, petrol and bread at three major outlets.
0
0
2
RT @toxnik: the oscar stage this year looks soooo easy to run onstage and disrupt everything.
0
7K
0
RT @GustavNilsonne: Do you want to do a postdoc in metascience? I am seeking 1-2 candidates whom I can host for a Marie Curie fellowship. Sā¦.
0
55
0