Peter PettoWrap up your APStats prep with a final push! Get tips from expert exam readers and connect with other sharp minds. You've got thisโand can be a contender! bit.ly/apreview2025 #APStats #StatEd #StatsChat #Statistics
Dr Mircea Zloteanu ๐บ๐๐If you have a 2x3 mixed factorial design, where the interaction is sig. can you fully analyse, report*, and plot this study using SPSS, JASP, or Jamovi? #rstats #statsodon #statschat #statsquestion #research
Yes (comment how)
Yes, but not fully
Not sure (just checking)
No (comment solution)
Poll ended at .
David Lawrence MillerPaul Blackwell (Sheffield) will be giving a workshop on "Modelling continuous-time capture-recapture data" on 5 December in Edinburgh. More details here: https://www.eventbrite.co.uk/e/modelling-continuous-time-capture-recapture-data-tickets-718374596757?aff=oddtdtcreator
still some places left!
Treat today at Edinburgh Uni stats seminar: Emiko Dupont (Bath) talking about her new work on spatial confounding (https://arxiv.org/abs/2309.16861) #statschat #statistics #mgcvchat
Demystifying Spatial Confounding
Spatial confounding is a fundamental issue in spatial regression models which arises because spatial random effects, included to approximate unmeasured spatial variation, are typically not independent of covariates in the model. This can lead to significant bias in covariate effect estimates. The problem is complex and has been the topic of extensive research with sometimes puzzling and seemingly contradictory results. Here, we develop a broad theoretical framework that brings mathematical clarity to the mechanisms of spatial confounding, providing explicit analytical expressions for the resulting bias. We see that the problem is directly linked to spatial smoothing and identify exactly how the size and occurrence of bias relate to the features of the spatial model as well as the underlying confounding scenario. Using our results, we can explain subtle and counter-intuitive behaviours. Finally, we propose a general approach for dealing with spatial confounding bias in practice, applicable for any spatial model specification. When a covariate has non-spatial information, we show that a general form of the so-called spatial+ method can be used to eliminate bias. When no such information is present, the situation is more challenging but, under the assumption of unconfounded high frequencies, we develop a procedure in which multiple capped versions of spatial+ are applied to assess the bias in this case. We illustrate our approach with an application to air temperature in Germany.
Lรฉo VarnetHello #statstodon! A reviewer asks me to perform a post-hoc #PowerAnalysis. I know this is generally not advised because if you replace the a priori effect size by the effect size measured in the experiment, this will introduce an erroneous relationship between the significance level of the test and the measured power.
โฆ but does that mean that there is no proper way of measuring power retrospectively? For example, if you refrain from using the measured effect size and instead simulate a range of โa prioriโ effect size unrelated to the results of the test, then the dependency of the power to the significance level should not happen?
#stats #statschat @lakens
โฆ but does that mean that there is no proper way of measuring power retrospectively? For example, if you refrain from using the measured effect size and instead simulate a range of โa prioriโ effect size unrelated to the results of the test, then the dependency of the power to the significance level should not happen?
#stats #statschat @lakens
#stats #statstodon #statschat #frequentist #NHST
Considering the concept of severe testing (Mayo), is there any point in planning for or running an hypothesis test if I instead determine my SESOI or predict a range of effects (e.g., 0.2<d<1.0)?
Shouldn't I just pre-register an equivalence test with those bounds, and avoid the "null hypothesis" test completely?
#statschat #statsdon #stats
Why would you multiple Cohen's d by sqrt(2) to get a "generic d"? ๐คทโโ๏ธ
See here:
.. the sample mean was 48.19, the expected age in the population was 50, so the difference would be 48.19 - 50 = -1.81. The standard deviation was 17.69, so Cohen's d becomes: d = -1.81 / 17.69 = 0.10. Cohen suggests to multiply this with the square root of 2, to get a generic d... (Cohen, 1988, p. 40).
#statschat #statsdon #stats #rant
While not totally accurate, do you think if we start saying "if you're wrong, the p-value is random" it would improve understanding of why they provide little info?
While not totally accurate, do you think if we start saying "if you're wrong, the p-value is random" it would improve understanding of why they provide little info?
Yes
No
Other (explain in comment)
Poll ended at .
#statschat #statsdon #statistics #stats
๐ฅ take: journals asking researchers to report p-values exactly (e.g., p = 0.032) has led to more people misinterpreting Neyman-Pearson Hypothesis Testing, and believing p-values are strength of evidence.
(instead of just reporting p < 0.05)
#statschat #statsdon #statistics #statistics
Do you know what Type M error and Type S error are?
(see Gelman & Carlin, 2014)
Yes
I've heard of it, but not sure
No
Poll ended at .