Kennisclip 1:
Both relate to the quality of a measurement instrument
Observed score (measured by the weighing scale) = true score (for example 70KG) + measurement error (random
error. difference between observed score and true score)
Reliability
- related to random measurement error
- If a weighing scale measures a different amount of KG every time and the true score is 70KG, the scale is unreliable
Measurement validity
- relates to systematic measurement error (for example weight indicated by measuring the height: it will
overestimate if a person is long and thin, it will underestimate if a person is small and fat)
- If a weighing scale measures 80KG ten times in a row and the true score is 70KG, the scale does not measure
what it is supposed to measure
A scale that measures ten times approximately 70 KG (69,8…70,2) is a reliable scale with a good measurement
validity
Kennisclip 2:
Internal validity of the experiment
XY
1. Poor measurement validity (how well you measure the concepts in your research question)
2. Faulty experimental design (can be solved by randomisation and a controlgroup)
Kennisclip 3:
- Take a sample of a target population
Design a sampling plan:
1. Define the target population (which belong to your population and which not)
2. Find or create a sampling frame (list) (over-coverage: there are people in your sampling frame that do not
belong there. Under-coverage: there are people missing in your sampling frame that should belong there)
3. Selecting a sampling method (random sampling method an d non-random sampling method)
4. Decide on the sample size
Kennisclip 4:
Cross-sectional design is used to describe a target population in terms of variables x and y in one moment in time
1. Poor measurement validity (interviewing students and ask if they committed plagiarism)
Both relate to the quality of a measurement instrument
Observed score (measured by the weighing scale) = true score (for example 70KG) + measurement error (random
error. difference between observed score and true score)
Reliability
- related to random measurement error
- If a weighing scale measures a different amount of KG every time and the true score is 70KG, the scale is unreliable
Measurement validity
- relates to systematic measurement error (for example weight indicated by measuring the height: it will
overestimate if a person is long and thin, it will underestimate if a person is small and fat)
- If a weighing scale measures 80KG ten times in a row and the true score is 70KG, the scale does not measure
what it is supposed to measure
A scale that measures ten times approximately 70 KG (69,8…70,2) is a reliable scale with a good measurement
validity
Kennisclip 2:
Internal validity of the experiment
XY
1. Poor measurement validity (how well you measure the concepts in your research question)
2. Faulty experimental design (can be solved by randomisation and a controlgroup)
Kennisclip 3:
- Take a sample of a target population
Design a sampling plan:
1. Define the target population (which belong to your population and which not)
2. Find or create a sampling frame (list) (over-coverage: there are people in your sampling frame that do not
belong there. Under-coverage: there are people missing in your sampling frame that should belong there)
3. Selecting a sampling method (random sampling method an d non-random sampling method)
4. Decide on the sample size
Kennisclip 4:
Cross-sectional design is used to describe a target population in terms of variables x and y in one moment in time
1. Poor measurement validity (interviewing students and ask if they committed plagiarism)
