t509blog

The 5 Whys: From Engineering to Education

"What problem(s) are we solving with blended learning, and what problem(s) are we creating?"

That is the question that prompted our small group discussion last week and ultimately created far more questions than answers. The discussion prompt was rooted, in part, in my own background in engineering and unrelenting desire to seek out the root cause of problems before attempting to solve them. If we immediately jump to solutions, we may end up catering to the symptoms and overlooking the underlying issues.

While this may seem obvious, it is precisely this type of problem solving that led Toyota to revolutionize automotive quality and along with it, the entire automotive industry (taking us from Mass Production to Lean Production). One of the most basic but effective methods pioneered by Sakichi Toyoda is the 5 Whys. The idea is to ask "Why?" enough times to discover the root cause (usually 5 is enough). Take a look at the example below.

Using the 5 Whys in an engineering context, informed by [1]

Instead of wasting time by constantly fixing parts or recalibrating the machine (symptoms), the real answer is to address the monotony of the operator's job (root cause). This could be accomplished by rotating the operators around the factory, involving them in the problem solving process, or maybe even giving them a raise.

Applying this same method to an example from education yields some interesting results.

Using the 5 Whys in an educational context, informed by [2]

This particular example highlights an issue alluded to in class that caught my attention. When teachers are very enthusiastic (aka pumped up) about a particular teaching methodology, it works better! In fact, this effect can overshadow the differences from one methodology to another. Perhaps this is also obvious, but maybe teacher enthusiasm shouldn't be something that is merely a measured byproduct of the teaching method. Could we maximize the effectiveness of new methodologies by designing them from the start to garner the most excitement possible from individual teachers?

Want to read more? Check out

[1] The Machine That Changed the World by Womack, Jones, and Roos.

[2] "Scaling Up: Evolving Innovations beyond Ideal Settings 
to Challenging Contexts of Practice" by Chris Dede

Putting IRT into Context

I found myself wondering what alternatives there are to item response theory (IRT). I soon discovered that IRT falls under the blanket of Pyschometrics along with a number of other theories, and so I thought it might be interesting to map them out.

Chart informed by https://en.wikipedia.org/wiki/Psychometrics

Once again, I wrote a python script to generate this chart. You can take a look at the code below:

import pygraphviz as pgv
import pdb
import sys

textfile = 'diagram3'
title = 'Psychometrics: IRT in context'

#Read through all the lines.
nodes = ['Pyschometrics',\
'Field of Study',\
'Psychological\nMeasurement',\
'Skills',\
'Knowledge',\
'Abilities',\
'Attitudes',\
'Personality\nTraits',\
'Educational\nAchievement',\
'Classical\nTest Theory',\
'Item Response Theory',\
'The Rasch\nModel',\
'Latent\nTrait Theory',\
'Strong True\nScore Theory',\
'Modern Mental\nTest Theory']
edges = [['Pyschometrics','Field of Study','is a              '],\
['Psychological\nMeasurement','Skills','evaluates      '],\
['Psychological\nMeasurement','Knowledge','evaluates      '],\
['Psychological\nMeasurement','Abilities','evaluates      '],\
['Psychological\nMeasurement','Personality\nTraits','evaluates      '],\
['Psychological\nMeasurement','Attitudes','evaluates      '],\
['Psychological\nMeasurement','Educational\nAchievement','evaluates   '],\
['Classical\nTest Theory','Psychological\nMeasurement','is a technique for                             '],\
['Item Response Theory','Psychological\nMeasurement','is a technique for   '],\
['The Rasch\nModel','Psychological\nMeasurement','is a technique for            '],\
['Item Response Theory','Latent\nTrait Theory','is also known as         '],\
['Item Response Theory','Strong True\nScore Theory','is also known as        '],\
['Item Response Theory','Modern Mental\nTest Theory','is also known as         '],\
['Item Response Theory','Classical\nTest Theory','outperforms'],\
['The Rasch\nModel','Item Response Theory','is a special case of     '],\
['Pyschometrics','Item Response Theory','encompasses               '],\
['Pyschometrics','Classical\nTest Theory','encompasses'],\
['Pyschometrics','The Rasch\nModel','encompasses              ']]


#Weed out all non unique edges
unique_edges = []
for x in edges:
    if x not in unique_edges:
        unique_edges.append(x)      

#Weed out all non unique nodes
unique_nodes = []
for x in nodes:
    if x not in unique_nodes:
        unique_nodes.append(x) 


print nodes
print edges

#make the graph
G=pgv.AGraph()
G=pgv.AGraph(strict=False,directed=True)

G.add_nodes_from(nodes)

for edge in edges:
    G.add_edge(edge[0],edge[1],label=edge[2])

sorted(G.edges(keys=False))

G.graph_attr['label']=title
G.graph_attr['labelloc']='t'
G.graph_attr['overlap']='false'
G.node_attr['shape']='oval'

s=G.string()
G.write(textfile+".dot")

G.layout(prog = 'dot')
G.draw(textfile+'.png')

First Post

Here's my first post! Check out Nicole's blog here!

Here's an interesting picture of Niagara falls that I took...

Here's an interesting picture of Niagara falls that I took...