puterleat

---
title: Initial analysis of vocabulary data
titlepage: true
make: worksheet
---


These initial analysis simply adjust for age. We don't try and adjust for SES yet – we can do that later if we're happy with these numbers and the approach.

I'm recommending we use a binomial model which estimates the probability of getting an item right given the covariates. We do this from number of correct items recorded, assuming 20 trials (the assumption is that if the item wasn't displayed to the child then they would have given an incorrect response).

puterleat

#!/bin/sh

# Some things taken from here
# https://github.com/mathiasbynens/dotfiles/blob/master/.osx

# Set the colours you can use
black='\033[0;30m'
white='\033[0;37m'
red='\033[0;31m'
green='\033[0;32m'

puterleat

    def _at_top_generator(self):
        """Introduce 5ms delay in readings and hysteresis for change in state"""
        
        pin = live_pins[self.name]['limit_top_pin']

        windowlen = 10
        window = deque(maxlen=windowlen)
        window.extend([True]*windowlen)
        state = True

puterleat

[1] -- Monday, January 19, 2015 -- 13:55:53
Exact - Proportions: Inequality, two independent groups (Fisher's exact test) 

Options:	Exact distribution

Analysis:	A priori: Compute required sample size 
Input:		Tail(s)                    	=	One
			Proportion p1              	=	.2
			Proportion p2              	=	.1
			α err prob                 	=	0.05

puterleat

#!/usr/local/bin/python

"""Script to select audio chunks from speech files."""

import operator
import os
import pydub
import itertools

puterleat

. xtmixed z_howstrong howmany || participant :

Performing EM optimization: 

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -530.88526  
Iteration 1:   log likelihood = -530.88526  

puterleat

2,100,100
10,500,500
5,550,600
5,600,700
5,650,800
3,750,1000
3,850,1500
3,800,1500
10,750,1500
10,700,1600

puterleat

[think different]

• Here to persuade you psychology is fun.
• Watch out — psychologists can play some funny tricks - you need to think for yourself...

[superhero]

Some psychology is about how amazing people are

puterleat

Protocol for prosody/therapy study

When sampling from files

• Coders rate odd and even patient numbers.
• We start with first and last session, then backfill other sessions when time allows
• Sessions are already broken into chunks of 5 minutes.

puterleat

     +---------------------------------------------------------------------------+
     |                 code                                   outcome   primar~e |
     |---------------------------------------------------------------------------|
  1. |           Dermen2011                             Drinking Days          1 |
  2. |           Dermen2011                   Drinks per drinking day          . |
  3. |     Fals-Stewart2005               Mean percent  days drinking          1 |
  4. |     Fals-Stewart2005                Mean percent days drinking          . |
  5. | Fals-Stewart2009 (1)          Mean percent days heavy drinking          1 |
     |---------------------------------------------------------------------------|
  6. | Fals-Stewart2009 (2)          Mean percent days heavy drinking          . |