Ch. 14 Stats Review and Data Cleaning
Measures of Central Tendency: Mean (average), Mean (middle value), Mode (most frequent response)
Tells us where the middle of the data is located
Measures of Dispersion: Range and Standard Deviation: Tells us how spread out the data are
Levels of Measurement Appropriate Measures of Central Tendency
Nominal Mode only
Ordinal Mode and/or Median
Interval/Ratio Mode, Median, and/or Mean
SPSS will give you anything you ask for, but the appropriate measure depends on what level of measurement the question is.
Mean: Sum of all data pieces divided by the number of respondents or number of data pieces
Median: 1) reorder the data from lowest to highest, 2) find the median position n+1 , find the median value
Mode: look for the number that shows up most frequently 2
Range: Highest Value - Lowest Value
Standard Deviation (sd): 1) fill out the table, then put the appropriate numbers into the equation
Data Base: 3 5 7 6 2 9 3
Mean: 3+5+7+6+2+9+3 / 7 = 35/ = 5.00
Mode: 3
Median: 1) 2 3 3 5 6 7 9
2) 7+1 / 2 = 8/2 = 4
3) Go to the 4th position on the reordered data base, Median Value=5
Range: 9-2 = 7
SD: n=7, mean=5.00
2 -3.00 9
3 -2.00 4
3 -2.00 4
5 0 0
6 1.00 1
7 2.00 4
9 4.00 16
0 38
Possible-Code Cleaning: If you had a response range from 1 to 5, but when you ran a frequency for that question you came up with a value of “7”, you would have an error in the data. Any value that is out of the original range.
Contingency Cleaning: You are checking the number of people who responded to either 1) the 2nd part of a contingency question to how many had answered “yes” to the 1st part. If 17 answered “yes”, then only 17 should have answered the 2nd part. OR 2) you check the number of people who answered one question based on how many answered another relevant question on the survey. If you asked “have you ever been pregnant” you could do a cross check on “gender” with “have you ever been pregnant” - you should find the “0” appears for the “Male” group.
|