Measurement Errors

Quick reference

• 00:05 Hi, I'm Ray Sheen, we talked about the importance of measurement system error.
• 00:09 It's now time to take a look at the factors
• 00:12 that can cause measurement system error.
• 00:14 So let's talk about what goes into a measurement.
• 00:19 When you record an observed measurement,
• 00:21 you're recording a combination of three effects.
• 00:24 On the one hand there is the true value of the item that is being measured, how big,
• 00:28 or how fast, or how hot it is.
• 00:30 However, there is also the measurement error associated with
• 00:34 the measurement system.
• 00:35 The uncertainty in the people and the equipment being used.
• 00:39 And there is a calibration standard of whatever measurement approach
• 00:42 you are using.
• 00:43 I wanna focus in on the measurement system error and its sources.
• 00:46 We're going to look at each of these in more detail.
• 00:50 But one thing I want you to realize is that there are many potential sources of
• 00:53 errors or error effects.
• 00:55 Some of these can be easily measured or analyzed, but
• 00:58 some will need a sophisticated statistical analysis.
• 01:01 The first one to discuss is accuracy, which is sometimes referred to as bias.
• 01:08 Accuracy is whether the average of the measurements of an item
• 01:11 agree with the true value of that item.
• 01:14 The question you are answering is whether or
• 01:16 not you are getting the right answer, at least on average.
• 01:20 If the answer to that question is yes, you are getting the right answer on average,
• 01:25 then we would say that the system is accurate.
• 01:27 What that means is that the difference between the actual value and the average
• 01:31 of the observed values as measured by the measurement system is very small.
• 01:36 When there are problems with accuracy,
• 01:38 we generally try to fix those by recalibrating the measurement system.
• 01:43 Now I'd like to talk about precision.
• 01:45 Accuracy dealt with the center point, on average, of the observed measurements.
• 01:49 Precision deals with the variability of those measurements.
• 01:52 In essence, the width of that bell shape curve.
• 01:55 It is determined by taking multiple,
• 01:57 independent measurements of the same item and recording the results.
• 02:01 This analysis is at the heart of most measurement system analyses.
• 02:05 And this error is often based upon both the design and use of the system.
• 02:09 The error can be divided into two categories based upon two
• 02:12 primary causes for error.
• 02:14 First is repeatability, this is looking at whether the same person,
• 02:18 measuring the same item with the same equipment and
• 02:20 in supposedly the same way, gets the same or different answers.
• 02:25 It's essentially looking at the variability of the equipment in
• 02:27 the process.
• 02:28 High repeatability would be a bell shaped curve that is very narrow,
• 02:32 low repeatability is a wide curve.
• 02:35 The other category is reproducibility.
• 02:37 Can different people using the same measurement system in measuring the same
• 02:41 item get the same result?
• 02:43 This is looking at person to person variability.
• 02:46 If there is high reproduceability each individual's bell shape curve is almost
• 02:49 identical with the others.
• 02:51 If low reproduceability, then each individual's curve is decidedly different.
• 02:57 Determining repeatability and
• 02:58 reproducibility is a statistical analysis known as a Gage R&R study.
• 03:03 It's a very practical and important one.
• 03:05 In our course on measurement system analysis,
• 03:07 we show you how to conduct the analysis and determine the results.
• 03:12 Next we will look at discrimination as a contributing factor to error.
• 03:15 Discrimination is the resolution of the measurement system.
• 03:18 That means a number of measurement intervals within the nominal range of
• 03:21 the item value.
• 03:22 It represents the ability of the system to determine
• 03:25 changes in the value of the item that is being measured.
• 03:29 Think of this as the difference between being able to choose only black or
• 03:33 white, versus different shades of grey.
• 03:36 At a minimum, for variable data we should have at least 10 gradations available in
• 03:40 the normal range of the item being measured.
• 03:42 And personally I would like to have 25 to 30 gradations.
• 03:45 In this illustration, you can see that the measurement scale on the left
• 03:50 only has three potential values for the distribution, 0, 1 or 2.
• 03:55 However, with the measurement scale on the right, we have much better discrimination,
• 03:59 since there are now 14 potential values in the range of the distribution.
• 04:03 The last two categories are stability and linearity,
• 04:06 both of which look at factors that change the accuracy or bias of the system.
• 04:11 The first is stability, and it looks at whether the bias value stays the same
• 04:15 while using the measurement system, or whether it drifts over time.
• 04:18 In our diagram, if we, again, let the gold bar represent the true value of the item,
• 04:23 the system starts with a negative bias in January.
• 04:26 It becomes more accurate, and, by March, there's virtually no bias.
• 04:30 And then it begins to drift even higher, and, in September,
• 04:34 the system inaccuracy is twice as high as it was low in January.
• 04:38 This could be due to wear and tear or
• 04:40 some aspect of the measurement system is being used up over time.
• 04:45 Linearity is also a bias shift, but this is not a shift over time.
• 04:49 Rather, this is a shift over the dynamic range of the measurement.
• 04:53 That means that the bias is one value when measuring small items and
• 04:57 another value when measuring large items.
• 05:00 If the bias stays the same over the full range of measurements,
• 05:03 then we say the linearity is good.
• 05:05 In this diagram, the bias is a small bias on the high side with small items and
• 05:10 a larger bias on the low side with large items, this is not good linearity.
• 05:16 Accuracy, discrimination, linearity and stability are measurement system
• 05:20 attributes that are normally based upon the design and maintenance of the system.
• 05:24 The precision element of repeatability and
• 05:26 reproducibility is based both on the design and use of the system.
• 05:33 Understanding your measurement system errors allows you to determine if
• 05:37 the overall measurement error is small enough that you can trust your data.
• 05:42 If the error is too large,
• 05:43 you probably need to change to a different measurement system.

Lesson notes are only available for subscribers.