Quality MGMT (sTATS) Flashcards

1
Q

What are sampling errors

A

when you sample and the smaple misleads you

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Type 1 error: producers risk

A

Sample says there is a problem, and there is no problem!!!

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Type 2 error:

A

sample says ok, and the sample is not okay!!

this is called consumer risk,

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

consumers risk and producer risk:

what is the action taken by producer

A

none

stop and investigate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

consumers risk and producer risk:
what is the actual popoulation

A

there is a porblme

there is is not problem

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

-> If a producer wants to avoid type 2 error/ consumer irsk, they want higher quality, but this will result in more type 1 errors

-> If a producer wants to avoid type 1 error/producer risk, they want low cost of stopping production, this will result in more type 2 errors

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

with a high quality strategy, what are you avoiding

are controls tight or loose

what type of error is increased

A

type 2 errors (consumer risk), controls are tights, more type 1 error (producer risk)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

with a low quality strategy, what are you avoiding

are controls tight or loose

what type of error is increased

A

type 1 error (producer risk), loose control, type 2 error (consumer risks)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

what are control charts

A

-graph used to study how a process changes over time
-eatimate the central value and likely range of variation of each statistic

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

what are control charts made up of!

A
  • central value
    -upper control limit
    -lower control limit
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

how to get the visual of a control chart
-green
-red

A

take a normal curve and turn it sideways

anything that is within 3 standard deviation on eiterh side is INHERENT RANDOMNESS VARIATION

anything that is outside the 3 standard deviations on either side in ASSIGNABLE/NON RANDOM VARIATION

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

slide 8.16

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

slide 8.25

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

what is N

A

the number of observations in the sample

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

what is M

A

the number of samples we take

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What does the green area in a control chart mean

A

this is the state of control
-> the process operating in its usual fashion
-variations areo nly caused by random fluctuations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

what does the red area mean

A

the process is out of control
-> not operating in its usual way
-> variaiton is not random and has a special cause

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

does out of control variation mean it is always a bad thing?

A

NO! merely indicate sthe process is not behaving as expected, given what is known

THIS DOESNT MEAN THAT OUT OF CONTROL=BAD ALWAYS

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

A control chart is a statistical process control

A

SPC! yesw

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

what are the two types of control charts

A

1) those for attributes
DATA WHICH COUNTS!
-> pchart

2) those for variables
DATA WHICH MEASURES!!!
-> x chart and r chart

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

P charts are used for what?

A

Data that is used to count (# of complaints)

ATTRIBUTES

22
Q

X charts and R charts are used for what

A

Data which measure
(anything in quantifiable units, cm, kg, ml, etc)

VARIABLES

23
Q

is THIS a attribute or variable?

errors in transactions
time spent waiting for teller
e transfer support calls

A

attribute
variable
attribute

24
Q

How to do control charts for attributes (p-chart)

A

1) set up the p-chart (find pbar, n, sd, UCL, LCL)
2) ifentify if the process is in control

25
How to set up the P-Chart
1) determine pbar -> the total number of defects from all samples/ (number of samples * sample size) 2) determine n 3) determine the standard deviation 4) identify the upper control limit 5) identify the lower control limit 6) draw this out, and identify if any points are out of control *if the LCL is negative, set LCL to 0
26
IF THE LCL IS CALCULATED AS A NEGATIVE NUMBER, THEN THE LCL IS
0
27
how do you do control charts for measurement by variables
USE TWO CHARTS IN TANDEM 1) chart for central tendency (x chart) 2) chart for variabiltiy (r chart)
28
why do we use two charts for measurement by variables?
because both the process variability and process average must be in control before the total process can be in control
29
What is the most common value of z?
3
30
what do we use the x chart to measure
the average/ or the cntral tendancy of the variables
31
what do we use the R chart to measure
the variability or the range of values
32
How to set up the X-Chart and the R chart
1) Find the value of n, or the number of observations in the sample -find the value of the grand mean (histroical average) -find the value of R bar (historical range) 2) solve for A2 using the chart 3) FOR THE X CHART: Calculate the sample mean (xbar), the upper limit, the lower limit 4) For the R chart: identify D3 and D4, solve for the UCL, Solve for the LCL NOTE: R=lARGEST VALUE-SMALLEST VALUE 5) check is process is in control for both of the charts by DRAWING UPPER AND LOWER CONTROL LIMITS!!! DO WE HAVE POINTS OUTSIDE THIS RANGE
33
note about xchart r chart details
The above approach, strictly speaking, requires that individual observations from the population follow a normal distribution. If this assumption is markedly violated, there are more appropriate procedures involving use of the sample std. dev. instead of its range.
34
Do we have some useless calculatioons for the proportion data?
no u dumb bitch this is the shit u plot fucking idito divide mistakes by n
35
Do we have some useless calculatioons for the variables data?
nou dumb bitch u plot this 1) get the sample avg 2) get the sample range (;argest-smalletst)
36
if one point is not inside the ucl and lcl, is process not in control
yes PROCESS IS NOT IN CONTROL AS A WHOLE
37
What are 4 additonal reasons to stop and investigates the data
1) if there is a trend in either direction (why is this happening) 2) if you get consistent points close to the UCL and LCL 3) if you have 5 consectuvite points above/below central line (no fluctating) 4) If you have erratic beaviour (not close to central line)
38
what are 2 other common issues in SPC
1) HOW often should you sample? -cost consideration, variability of process, cost of quality faults (how bad is it if quality is poor? think pencil vs pacemakers) 2) WHERE to monitor/control in a multistage process - before "costly" stagesi n process -at the end to ensure satisfacitons for customers -at histroically unreliable stages - near beginning to isolate supplier problems
39
why cant we monitor process at every stage
too expensive and slows it odwn
40
what is process capability
measures whether or not output will routienly meet the desin specificaitons WHEN THE PROCESS IS IN CONTROL
41
what are specification limits on process capability
The UPPER SPECIFICAITON LIMITS and LOWER SPECIFICAITON LIMITS (USL AND LSL) are set externally and not affected by process immprovement and sampling LIKE CSE helmets and NTS crash ratings
42
what are specifiaiton limits for process capcabiltiy measured by
relationship of variation of INDIVIDUAL VALUES of the process with the USL and LSL -> think that in this you focus on the individual procut- these must meet the specifications!
43
4 types of process capability charts: 1) if normal graph is touching both usl and lsl 2) if normal graph is narrow around the mean and not touching usl and lsl 3) if edges go past lsl 4) if the grand mean has shifted so a litttle bit is outside usl
yes-JUST BARELY YES EXCELLENT NO NO-MEAN HAS MOVED
44
SLIDE 8.56
45
How to do process capabiltiy examples
1) use the Zusl formula or the Zlsl formula!!! 2) Identify the USL or LSL, the given sample mean (x bar), and the sd 3) use z table to convert z score into a probability 4) read quesiton critically: how much probability is allowed on either tail? DRAW OUT THE NORMAL CURVE!! USE THE CORRECT TAIL OF THE PROBABILITY!!
46
how is six sigma used in assesing process capability?
once you calcuate the defectiveness, see if you can get it to be only 3 per million defective of 0.0000003 defective this means that a prcoess is EXTREMELY CAPABLE!!! AND ABIDES BY 6 SIGMS
47
how to visualize 6 sigma
draw a normal chart take 6 sd jumps from core yhe prob is 99.9999998% effective and 0.0000002% ineffective!! if this is the effeciteness thatn u won!!! you abide by 6 sigma
48
Slide Notes: Out of Control does not mean it is bad/non-functional/sometimes these defects can still be functional and okay to use, but just outside of quality management Polygraphs lie detectors are a good control charts
49
when we use USL, what area do we care about
area to the left of USL, "15 M OR LESS"
50
WHEN we use LSL whta area do we care about
area to the right of LSL
51