Obs. and measurement

Question 1

Observation

Answer 1

NARROW VIEW: You observe only what you can detect with your senses.

• Direct observation: by looking at, hearing, etc. without any aid
• Aided direct observation: looking at, hearing, etc. possible only with aids

WIDE VIEW

• Indirect observation: you observe also when you detect the effect of an event, not the event itself
• In order to associate even of interest with observed effect, we need theoretical knowledge
• Observation, at least in the wide sense, is theory-dependent

Question 2

Confirmation bias (2 types!)

Answer 2

COLLECTION of evidence:
searching for evidence that could confirm your hypothesis
INTERPRETATION of evidence:
The observer ”sees” what he or she expects to see

Question 3

Hawthorne effect

Answer 3

People have a tendency to change their behaviour when they know they are being studied.(type of observer influence)

Question 4

Single blinding

Answer 4

The subject does not know whether he or she is getting a treatment or not.

Question 5

Double blinding

Answer 5

The experimenter does not know which subject is getting the treatment. Reduces the effect of observer expectations.

Question 6

Operationalization

Answer 6

• A way of translating an idea into something measurable or concrete.
• The measurable or concrete thing that represents an idea.
• To provide a way of linking it to observable properties.

Question 7

Construct validity

Answer 7

(aka the accuracy)of a measurement system is the degree of closeness of measurements of a quantity to that quantity’s true value.

Question 8

Convergent validity

Answer 8

different ways of measuring the same intended concept should give rise to similar results.

Question 9

Divergent validity

Answer 9

different ways of measuring different intended concepts should give rise to dissimilar results.

Question 10

difference between precision and accuracy?

Answer 10

~~in regard to a measurement system~~

A: (“construct validity”) the degree of closeness of measurement of X to X’s true value.

P: (“reliability”) the degree to which repeated measurements under unchanged conditions show the same results.

(think target pictures, precise, not accurate off to the side. accurate but not precise all around)

Question 11

stability of a measurement system?

Answer 11

The degree to which similar measurements repeated over extended periods of time show the same results

Question 12

Scale types (5)

Answer 12

(connected to representation by numbers):

1. Nominal scale: non-identity (assigning samples identifying numbers)
2. Ordinal scale: positive monotone transformation (Mohs scale for hardness)
3. Interval scale: positive linear transformation (temperature in Celsius)
4. Ratio/quotient scale: positive scalar transformation (length, volume, temperature in Kelvin)
5. Absolute scale: unique (counting objects, probability)

Question 13

Random errors

Answer 13

Inherently unpredictable and have null expected value – i.e. they are scattered about the true value.

• They can be reduced by repeating measurements and averaging over outcomes, or by trying an alternative operationalization that might have lower random errors.
• Reducing random errors increases precision, but not accuracy.

Question 14

Calibration of an instrument

Answer 14

…evaluating and adjusting the PRECISION AND ACCURACY of measurement equipment.

…intended to eliminate or reduce bias in an instrument’s readings over a range for all continuous values.

Question 15

Measurement

Answer 15

• Assignment of a number to a characteristic of an object/event (which can be compared with other objects or events)
• Definition of concept to be measured
• Representation by numbers

Question 16

Systematic errors

Answer 16

..predictable and typically constant or proportional to the true value.

• If the cause of the systematic error can be identified, then it usually can be eliminated.
• Reducing systematic errors increases accuracy, but not precision.