To ensure that your results are valid, consider the following
guidelines when you collect data, perform the analysis, and interpret your
- To analyze a mean, median, sum, variance, or standard deviation, the
data must be numeric
- You must have continuous data, such as the weights of packages, or
discrete data, such as the number of complaints. If you have binary data, such
as yes/no or pass/fail, analyze a proportion.
- To analyze a proportion, the data can contain only two categories, such
as pass/fail and 1/0
- If you have continuous or discrete data, such as length, weight, or
number of complaints, select a function for continuous or discrete data, such
as the mean.
- The sample data should be selected randomly
- In statistics, random samples are used to make generalizations, or
inferences, about a population. If your data are not collected randomly, your
results may not represent the population. For more information, go to
in samples of data.
- Each observation should be independent from all other observations
- For observations to be independent, the value of a particular
observation does not depend on any previous observation. If your observations
are not independent, your results may not be valid. For more information, go to
are dependent and independent samples different?.
- The sample size should not be small
- If your sample size is small, the resampling results may be
unreliable. To ensure that your results are valid, collect a medium to large
sample. An adequate sample size depends on the characteristics of the data. Use
the histogram to determine whether your sample size is large enough.