Chapter 9 - Better Measurements

Question 1

What is Calibration?

Answer 1

Measuring something for which you already know the answer to, to see how accurate your measurement tool is.

Question 2

What is the objective with Calibration?

Answer 2

To lower bias and avoid inaccurate results.

Question 3

How can you reduce the problem of a fault occurring in your measurements?

Answer 3

By saving and measuring them immediately one after the other at the end of the 14 days.

Question 4

Define Randomization:

Answer 4

The order in which you take measurements (or the timing of measurements) might need to be randomized to avoid bias

Question 5

The __________ of your samples might change over time.

Answer 5

weight

Question 6

Your measurement tool might ______ in accuracy over time.

Answer 6

drift

Question 7

Define Accuracy:

Answer 7

Accuracy measures how close results are to the true or known value.

Question 8

Define Precision:

Answer 8

Precision measures how close results are to one another.

Question 9

Using the race times example, why is using a digital stopwatch more precise than an analogue watch?

Answer 9

Because the analogue requires some interpretation.

Question 10

If you have to use a tool that is less precise, how can you ensure consistency in your measurements?

Answer 10

You can set up rules for interpretation that any observer must follow.

Question 11

In the race times example, what happens if both of your watches are running fast?

Answer 11

It is prone to introducing bias or inaccuracy

Question 12

In the race-time example, how can calibration assist if both of your stop-watches are running fast?

Answer 12

You can correct the measurements after the fact, by detecting and quantifying the bias.

Question 13

Why is it important to randomize the order or sequences of subjects to the observer?

Answer 13

To minimize bias upon the measurements of the treatment group.

Question 14

What is Intra-observer Variability?

Answer 14

Variability in measurements due to a single observer “drifting” in how they measure something.

Question 15

What is one way of avoiding Intra-observer Variability?

Answer 15

By using a quantitative measurement and then categorizing the measurements afterwards

Question 16

What are two ways of minimizing Intra-Observer Variability?

Answer 16

• Regularly calibrating yourself to your own ranking system.
• Randomizing the order in which you rank individuals from different treatment groups.

Question 17

Define Inter-observer Variability:

Answer 17

Variability in measurements due to two or more observers differing in how they measure something.

Question 18

How can you avoid Inter-observer Variability?

Answer 18

By testing each observer’s measurements against known values.

Question 19

What is a Repeatability Study?

Answer 19

When measuring random samples several times and comparing their scores to identify any observer drift.

Question 20

What do you do if no known values are available when measuring Inter-observer Variability?

Answer 20

You can test them against an expert’s measurements.

Question 21

To mitigate Inter-observer variability, avoid situations in which ______________________________.

Answer 21

One observer measures one whole treatment group and another measures another group.

Question 22

What are three ways of mitigating Inter-observer Variability?

Answer 22

• Keep track of who the observer was
• Less subjective measurement
• Regular Calibration between observers

Question 23

The more _____________ you have in your design, the more chance for error/bias.

Answer 23

categories

Question 24

With multiple observers, _____________ is needed to ensure each collector categorizes individuals consistently.

Answer 24

collaboration

Question 25

Which is easier to combat, Inter-observer or Intra-observer Variability, and why?

Answer 25

Inter-observer Variability because when two people disagree on scoring of the same thing, then you can have a discussion on why this has occurred and how to prevent it.

Question 26

What is often the root of inter and intra-observer variability?

Answer 26

Assigning individuals to categories

Question 27

What is an example of an observer effect?

Answer 27

Accidentally spooking a fish while trying to detect behavioural response to a model predator.

Question 28

What is the Observer Effect?

Answer 28

The fact that observing a situation or phenomenon necessarily changes it.

Question 29

What are the five key things to note when recording data?

Answer 29

1. Avoid recording too much data at once.
2. Avoid recording data under conditions that can lower the data quality.
3. Be careful with data codes (well-defined)
4. Ensure data collection protocols are clearly written and note changes made in the field.
5. Keep more than one copy of data in multiple locations.