7

excess significance bias

underpowered studies

no replication work done

data analyses , protocols not publicly available

more likely to report statistically significant results than are studies that report a nonsignificant conclusion, especially for smaller studies

when results from scientific research are deliberately not fully or accurately reported, in order to suppress negative or undesirable findings

violation of impartiality in the evaluation of a submission

making up data or results and recording or reporting them

salmon study

aim to show that some scientists are too excited baout research methods

the results might not be reliable as they depend on the measure used

salmon was scaned in fMRI and the images showed that it does indeed have a correlation even withour a possibility at peprspective taking

study provides evidence of individuals with ADHD have altered smaller brains

however ther as some important shortomings on the part of the reserchers

conclusion due to "effect size"

mean IQ in ADHD indivduals was at least equal or higher than control group

however these findings were not published

in order to see table of IQ results 30 odlas had to be paid

why did they not publish these results and instead they published some false informatio on brain size

after people asked they renewd data on IQ to show no signifiance

if we assume mindrain are related, we will find differences between people with ADHD and control groups

need to overcome the nature/nurture dichotamy

3 control groups

1. scientists
2. PhD students
3. control

given a study to examine with different 4 explanations

2 good ( neuroscientific/non neuroscientific)

2 bad ( neuroscientific / non neuroscientific)

even if the neuroscientific example was not relevant at all to what they were talking, the non espert group considered it more "correct" than the explanation without

articles with images had higher ratings of scientific reasoning considered more true than the ones without

reserch lies in the presuasive power of the actual brain images

influentia beacuse they provide physical basis

people like simple efforts but more pictures does not equal more credibility

there is a difference between photo ( photo of what actually there is e.g. a tree) and image ( capute of what is happening which does not alwys show exctly what we see)

we do not actually undertand brian images as they have a lot of characteristics

1. do we need to bother people with the intricacies of scientific research?
2. do we need to communicate our teoretical assumpotions and metodological doubts and uncertinities
3. if there is "dirty laudry": should we go public ? what about funding ?
4. what can we promise ? does science need marketing?

how we interpert data is alweys be in terms of what our own theoretical perspective is

a lot of times 'data consistent' with hypothesis → but is it also consistent with null hupothesis?

it is not enough for the data to be 'consistent' with the hypothesis

we need to asses probability of the data conditional on the null hypothesis

( in this case : mechanisms responsible for face perception aren't specialized or localized)

he publish or perish culture

pressure to publish brought different scientists to publish more with less quality

"Dr. Max published 138 papers in a year ( one every two days )

his PhD students would write his name as well on the papers, not actually good studies

therfor now we need to say which participant did which part of the papers we publicate

reviewer process has been installed as well

pape written as humorous outake on the fate that awaits scientists that will sin agains best practice

each of the 9 layers is a worst and worst action against publication with last being falsification and the top being the " limbo ": awarding or accepting the ones who do worst

found 55 cause of fraud and 10 highly probable fraud since 2004

this cause damage in multiple sites :

co

faith in science

the reputation of the universities that took part in such publications

1. Fabrication : entering fictiotious data
2. Falsification : falsyfing data and/or rejecting research results obtained
3. Plagiarism : plagiarise publications and respults of others
4. Deliberatly ignoring or not acknowledging contributions of other authors
5. Falsely writing name as author
6. Intentionally making incorrect use of methods and/or intentionally interpreting results incorectly
7. commiting attributable inaccuracie whn carrying out research
8. allowing and conciling misconduct to collegues

1. Scrupulousness (unaffected by mounting pressure to achieve )
2. Reliability
3. Verifiability
4. Impartialioty
5. Independece

worries about impact of the replciation crisis and methodological problems

anti

common misconception that science is "derivated form facts"

1. facts are dicetly given to careful, unprejudiced observers via the senses
2. facts are prior and independent of theory
3. facts constitute a reliable foundation for scientific knowledge

observation is not determined soley by the object view

the skils and observer has are not identicalto a novice , this changes perveption

e.g. if not trined to see a tumor you can look at an X ray and not see anyting

if the Earth moves and you drop a stone form the top of a tower it should strike the ground some distance from the tower

everyday obervations shows that it is wrong

we need theory mediated observations to prove

need to distinguish fascts as statemnts about states of afairs FROM states of affairs and FROM percetions of those facts

botanish will go in a garden and see a lot of things that we cannot

need **language and expirince** to be able to do so

scientific observations are based on expirince, knowledge and expectations of the observer

there are a lot of theories that are non observable entitis, they can only be explained and predicted

acknowledgeing the role of the theory is not problematic

=theory can be logically derived from facts

problematic to claim that scientific knowledge is derived form facts

1. Metal X expands in heath
2. Metal Y expands in heath
3. Metal Z expands in heath

= all metals expand in heat = not a valid argument ( premisis is true but conclusion is not )

1. lange nr of A's have been observed under
2. a wide vareity of conditions and if all of those A's
3. without exception posses the porperty B, then all A'd have the property B

1. cannot account for unobservable entities postulated by contemporary science
2. observation emasures is imprecise and inexact
3. we cannot justify the induction prinicple

falsification