pgomez1dpu

A session in HDDM

Needed as in the DEMO

inline

pgomez1dpu

---
title: "OTEs and Effectiveness"
author: "Pablo Gomez (PSY)"
date: "February 8, 2016"
output: pdf_document
---


The first step is to measure whatever can be easily measured. This is OK as far as it goes. The second step is to disregard that which can't be easily measured or to give it an arbitrary quantitative value. This is artificial and misleading. The third step is to presume that what can't be measured easily really isn't important. This is blindness. The fourth step is to say that what can't be easily measured really doesn't exist. This is suicide.

pgomez1dpu

https://gist.github.com/pgomez1dpu/9b0c8ee1a182ca8f8b71#q7-


https://gist.github.com/pgomez1dpu/9b0c8ee1a182ca8f8b71#q8-


https://gist.github.com/pgomez1dpu/ecdb2c8752f280629ef0

pgomez1dpu

Condition Error RT Correct RT

SP-SP 3851(472) 1833(70)

DL-PH 5038(779) 1955(62)

GC-GX 3028(285) 2133(59)

TZ-ZT 2596(208) 2336(107)

pgomez1dpu

Sanjay Srivastava: Is there any way under standard, textbook NHST to reliably get .04<p<.05 more than p<.01? I dont think there is

nsims  <- 100000
effect <- .1
result <- array(dim=nsims)

for (i in 1:nsims){

pgomez1dpu

% This file is a LaTex file for the "beamer" package to create slides for a talk or presentation


% MODIFIED by Pablo Gomez 5/20/2015

\documentclass{beamer}


\mode<presentation>
{

pgomez1dpu

#Your turn. LME 1

Suppose that you have an experiment in which you are presenting faces in two different conditions: in a group, and by themselves. You want to test if people look more attractive in a crowd.

look here

How would you analize this data?

Your turn. LME 2

pgomez1dpu

Previously, in this class...

filename <- "http://www.uvm.edu/~dhowell/methods8/DataFiles/Logistic.dat"
logistic <- read.table(filename, header=T)

log.recode <- logistic
log.recode[logistic$outcome == 2, 2] <- 0
plot(log.recode$survrate, jitter(log.recode$outcome,.1))

pgomez1dpu

Your Turn (BayesPrimer.1)

Use one paragraph to answer this question: Can a p = . 05 be more consistent with the H₀ than with the H₁ hypothesis? Let's suppose that you run a regression, and the p value for a coefficient is .05. How sure are you that the H₀ is not true?

Your Turn (BayesPrimer.2)

Related to the previous question, if the coeficient was .25 in y = .25 * x + Error how likely is it that you would get a p=.05? More likely than if the coeficient was 0?

Your Turn (BayesPrimer.3)

pgomez1dpu

This follows Howell's Statistical Methods for Psychology, 8th ed

Read page 556

Get the data that the book refers to a survrate.dat (it really is called Logistic.dat )

filename <- "http://www.uvm.edu/~dhowell/methods8/DataFiles/Logistic.dat"
logistic <- read.table(filename, header=T)