Interpreting statistical information
Statistics are raw data and must be processed to create meaningful informationAverages: a measure of data's central tendency'
- Arithmetic mean (x) = add all items and divide by number in set
- Median = middle value when you arrange data in ascending order (if even number, mid-way between the two)
- Mode = most frequently occurring value
classes. Count the number of items of data that fall into each class (the class frequency).Compare relative frequencies or show trends by comparing the same class over time.
Cumulative frequency distribution uses ceilings instead of ranges to define classes e.g. under 9
Trends: underlying movement in time series over the long term. (Time series: set of data recorded at
intervals over period of time.)Patterns that can be observes in business time series:
- Trends - sales revenue may show positive trend, a negative trend might be staff turnover
- Seasonal patterns/variations - peaks and troughs at same times of each successive year, perhaps showing a seasonal pattern. There may also be a general trend, are the peaks and troughs getting higher each year?
- Cyclical variations - reflect the larger pattern of swings in the economy (boom & bust)
- Random variations - irregular movements reflecting unpredictable factors. They may not affect overall trend.
- What's the argument or story? - think as if writing press release. What's newsworthy? What information can you use?
- What comparisons are suggested? - translate into pie chart or bar graph to see possibilities
- What trends or correlations can be observed? - Useful in summarising importance of data and indicating if anything needs to be done. A trend suggests change. A correlation suggest is you manipulate one variable you can change another. Mentally see results in a scattergraph.
- How reliable and meaningful is the data? - Are anomalies or surprises accounted for by age, sample size, type of questions asked?
- Are you interpreting the data correctly?
No comments:
Post a Comment