Miller’s Law

1 / 35

# Miller’s Law - PowerPoint PPT Presentation

Miller’s Law. “In order to understand what another person is saying, you must assume it is true and try to imagine what it might be true of.” George Miller. Models of Experts Outpredict the original . Internists diagnosing disease College admissions committees Airplane autopilots Why?.

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.

## Miller’s Law

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Miller’s Law
• “In order to understand what another person is saying, you must assume it is true and try to imagine what it might be true of.” George Miller
Models of Experts Outpredict the original
• Internists diagnosing disease
• Airplane autopilots
• Why?
Why Models Work Better
• Explicit Criteria
• Consistent application Valid comparisons
• Reduce random error
• Eliminate irrelevant criteria
• Eliminate prejudice based on irrelevant data
• They follow from the data
• Where does the data come from?
Statistical Process Control (SPC)
• Uses statistics & control charts to identify when to adjust process.
• Involves:
• Creating standards (upper & lower limits).
• Measuring sample output (e.g. mean weight).
• Taking corrective action (if necessary).
• Done while product is being produced.
Outline
• Statistical Process Control (SPC).
• Mean chartsor X-Charts.
• Range chart or R-Charts.
• Control charts for attributes.
• P charts--% defective
• C charts—number of defects per piece
• Acceptance Sampling.
Statistical Process Control (SPC)
• Statistical technique to identify when non-random variation is present in a process.
• All processes are subject to variability.
• Natural causes: Random variations.
• Assignable causes: Correctable problems.
• Machine wear, unskilled workers, poor materials.
• Uses process control charts.
Control Chart Types

Control

Categorical or Discrete Numerical Data

Charts

Continuous Numerical Data

Variables

Attributes

Charts

Charts

R

P

C

X

Chart

Chart

Chart

Chart

Quality Characteristics

Variables

Attributes

• Characteristics for which you focus on defects.
• Categorical or discrete values.
• # of defects.
• Characteristics that you measure, e.g., weight, length.
• Continuous values.

Plot of Sample Data Over Time

80

Upper control limit

60

Sample Value

40

20

Lower control limit

0

1

5

9

13

17

21

Time

Process Control Charts
Control Charts
• Process is not in control if:
• Sample is not between upper and lower control limits.
• A non-random pattern is present, even when between upper and lower control limits.
• Based on sample being normally distributed.
X Chart
• Shows sample means over time.
• Monitors process average.
• Example: Weigh samples of coffee.
• Collect many samples, each of n bags.
• Sample size = n.
• Compute mean and range for each sample.
• Compute upper and lower control limits (UCL, LCL).
• Plot sample means and control limits.
Distribution of Sample Means

Standard deviation of the sample means

(mean)

Central Limit Theorem

As sample size gets large enough,

distribution of meanvalues becomes approximately normal for any population distribution.

Central Limit Theorem

sample mean at time i

 = known process standard deviation

X Chart - Example 1

Each sample is 4 measurements.

Process mean is 5 lbs.

Process standard deviation is 0.1 lbs.

Determine 3σ control limits.

R Chart
• Shows sample ranges over time.
• Sample range = largest - smallest value in sample.
• Monitors process variability.
• Example: Weigh samples of coffee.
• Collect many samples, each of n bags.
• Sample size = n.
• Compute range for each sample & average range.
• Compute upper and lower control limits (UCL, LCL).
• Plot sample ranges and control limits.
p Chart
• Attributes control chart.
• Shows % of nonconforming items.
• Example: Count # defective chairs & divide by total chairs inspected.
• Chair is either defective or not defective.
c Chart
• Attributes control chart.
• Shows number of defects in a unit.
• Unit may be chair, steel sheet, car, etc.
• Size of unit must be constant.
• Example: Count # defects (scratches, chips etc.) in each chair of a sample of 100 chairs.
Acceptance Sampling
• Quality testing for incoming materials or finished goods.
• Purchased material & components.
• Final products.
• Procedure:
• Take one or more samples at random from a lot (shipment) of items.
• Inspect each of the items in the sample.
• Decide whether to reject the whole lot based on the inspection results.
TQM - Total Quality Management
• Encompasses entire organization from supplier to customer.
• Commitment by management to a continuing company-wide drive toward excellence in all aspects of products and services that are important to the customer.
Three Key Figures
• W. Edwards Deming
• Management & all employees have responsibility for quality.
• 14 points.
• Deming Prize in Japan.
• Joseph Juran
• Focus on customer.
• Continuous improvement and teams.
• Philip Crosby
• Quality is free!
• Cost of poor quality is underestimated.
Costs of Quality
• Internal failure costs.
• Scrap and rework.
• Downtime.
• Safety stock inventory.
• Overtime.
• External failure costs.
• Complaint handling and replacement.
• Warranties.
• Liability.
• Loss of goodwill.
Why TQM Fails
• Lack of commitment by top management
• Focusing on specific techniques rather than on the system
• Not obtaining employee buy-in and participation
• Program stops with training
• Expecting immediate results rather than long-term payoff
• Forcing the organization to adopt methods that aren't productive or compatible with its production system and personnel
• from Martinich, Production and Operations Management
Customer-focused Quality Management:

We treat our employees like dirt

and pass the savings on to you.

Taken in isolation, each step is valid and acceptable...

A = B

A2 = AB

A2 - B2 = AB - B2

(A + B) (A - B) = (A - B) B

(A + B) (A - B) = (A - B) B (A - B) (A - B)

(A + B) = B

A + A = A

2A = A

2 = 1

But the overall result is absurd.

Total Quality Management---
• Focus on the Long Term best average result rather than immediate short-term outcome.
• Emphasize process rather than single result.
• Design quality into the process rather than testing defects out of the product.
• Aim for zero defects through continuous improvement.
• Base vendor decisions on relationship and statistical evidence of quality rather than price.
• Buy value rather than price.
• Reduce perception of personal risk in decision making.
• Drive out fear.
• Foster rational laziness.
• Let People do the things that are important
• and they will seek out the important things to do.
• Explicit goals and criteria for success
• Consistent best bet decisions
• Efficiency with resources
• Freedom from Fear
• Concern for welfare of the organization
• Global view of the organization
• People
• Geography
• Time

• Myopia
• Personal expediency
• Fear of blame
• Avoidance of perceived personal risk
• Disregard for long term welfare and lack of concern for others.
Most people are busy--
• Being concerned about personal risk
• Trying to avoid failure
• Afraid of being blamed for occasional misfortunes
• Don’t want to take responsibility

Some people are too busy--

Some people are too busy--
• Being managers