14- Error and Confidence Interval

Question 1

What does Error mean?

Answer 1

The Difference between our estimate of a parameter and the true parameter

Question 2

What is Sampling Error?

Answer 2

The error associated with the sample statistics we calculate (mean , proportion, probability)

Question 3

What is Systematic Error?

Answer 3

A non-random error associated with measurement, instruments, sampling method and other factors (includes measurement error and methodological error)

Question 4

Four Characteristics of Measurement Error

Answer 4

1) Accuracy limit- Granularity of the reporting interval. 2) Range limits- Detectable only in a certain range. 3) Response Lag- time between the measurement and the result. 4) Level of Detection- Limit at which detectability is possible

Question 5

Three Characteristics of Methodological Error

Answer 5

1) Biased Sampling- sample of a subset of population is a biased representation of the whole population 2) Incorrect Statistical Measure 3) Miscalculation- Error in programming code or in the manual calculation of statistics

Question 6

Can methodological errors be eliminated? If yes, how?

Answer 6

Yes, through choosing the appropriate elimination method

Question 7

What type of systematic error can be unavoidable?

Answer 7

Measurement Errors, but we can usually take them into account to ensure that our findings are valid

Question 8

What is Measurement or Instrumental Error?

Answer 8

It is an error related to the tool we are using to analyze data

Question 9

What causes Methodological Error?

Answer 9

Sampling biases, poor choices about statistic and miscalculations

Question 10

How do we represent sampling error?

Answer 10

Using point estimators as they represent an unknown population parameter

Question 11

What are Interval Estimators?

Answer 11

A range/interval estimator is calculated from a sample and provides a range of plausible values, has upper and lower bounds, does not account for instrument errors, biases, or other methodological errors

Question 12

What are Confidence Intervals?

Answer 12

tells us about a range that is likely to include the true unknown population parameter with some level of certainty

Question 13

Confidence Interval Relationship

Answer 13

Smaller sample sizes result in larger confidence intervals, larger confidence levels results in larger confidence intervals, larger standard deviations result in larger confidence intervals

Question 14

Purpose of Confidence intervals

Answer 14

To estimate the interval within which a population parameter is expected to fall at a certain confidence level (Narrower the interval the better)