Instats

Instats is the world's premier research methods and statistics training platform and AI software provider for research intensive organizations.

[Instats @instats_]

Livestreaming Dec 15, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/5B6B3gyeqD

[Instats @instats_]

Researchers risk losing independent judgment when delegating to AI. Draft your own position before seeking AI critique, and require approval before an assistant acts.
#AgenticAI #ResearchMethods

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Nov 16, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/7Y5zdNz83v

[Instats @instats_]

Dark energy and curvature can be hard to disentangle. Fit flat and curved models to the same data, then inspect parameter degeneracies before claiming evidence.

#Cosmology #Statistics

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Feb 18, 2027 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/9FPAZGydNU

[Instats @instats_]

Is a transfer student an attriter? Define persistence using your theory, then distinguish transfers, stop-outs and departures when they represent different processes.
#HigherEd #ResearchMethods

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Dec 7, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/C1IMPyEUAK

[Instats @instats_]

Researchers often mistake animal behavior for evidence of experience. In ELAN, annotate observable actions separately from interpretations; record the evidence for each claim.
#Biosemiotics #ELAN

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Dec 3, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/b2J0Riofuf

[Instats @instats_]

Null results in survey experiments can be hard to interpret. Use a power analysis to plan sample size, and pilot your treatment to check that respondents understand it.

#SurveyExperiments #CausalInference

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Dec 2, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/8iL9lstTM9

[Instats @instats_]

Vaccine effects are hard to estimate when people influence each other. Map social ties and shared places; consider effects on recipients and their contacts.
#CausalInference #Epidemiology

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Nov 16, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/d0TBtrKMLs

[Instats @instats_]

Genomic models can overfit scarce data. Fit preprocessing on the training set only, and tune hyperparameters on a separate validation set.
#DeepLearning #Genomics

Full details and registration in the first reply.

[Instats @instats_]

Livestreaming Dec 1, 2026 (UTC), with the recording available afterwards.

Full details and registration: https://t.co/jLvpKMKpcU

[Instats @instats_]

Separation can make logistic regression estimates diverge. Check for extreme coefficients and standard errors, even if the model converges; consider Firth’s penalized likelihood.
#LogisticRegression #Biostatistics

Full details and registration in the first reply.

[Instats @instats_]

Available on demand, at your own pace.

Full details and registration: https://t.co/z3JSi2Yl4F

[Instats @instats_]

LLM-derived variables can introduce measurement error into causal estimates. Compare outputs with human-coded examples, then check whether errors differ by treatment group.
#CausalInference #LLM

Full details and registration in the first reply.

[Instats @instats_]

Available on demand, at your own pace.

Full details and registration: https://t.co/xYCpnfCAYF

[Instats @instats_]

Genomic clusters can be hard to define. Standardize features before k-means; try a Gaussian mixture model for probabilistic cluster assignments.
#Genomics #MachineLearning

Full details and registration in the first reply.