Descriptive Statistics

Basic Math and Key Measures

• There are 5 easy to compute measures to derive meaning in most data analysis situations.
• percentage
• average
• median
• mode
• standard deviation

Percentage

• The use of percentages for the analysis of survey data works exactly as you'd expect. Divide the number of people who give a specific response by the total number of people answering the question.
• Once you calculate the percentages for each type of response, you can chart them
• The total of the percentages should add up to 100%

Averages

• You probably remember this from your classes. Your class average summarizes your performance in the class.
• Averages are useful to summarize responses to a particular survey question
• Divide the total number of points for the question by the total number of people answering the question

Median and Mode

• the median is the number that appears in the middle of an ordered set of data
• (lowest to highest)
• useful when an average is skewed by a very high/low response
• the mode is the number that occurs most often
• useful to find the most frequent response

Standard Deviation

• Dispersion refers to how spread out/clustered a set of scores is around the mean
• Standard deviation is a way to understand and compare distribution of scores without having to actually plot and examine each and every score
• When comparing standard deviations, any particular standard deviation is a reflection of the underlying scale
• Standard deviations cannot be compared when scales have different numbers of options
• Look at page 403 for the standard deviation formula
• This is what SPSS is used for...

MAKE SURE YOU HAVE GOOD DATA

• Clean up your data and fact check it
• check responses to see if the are responses that are:
• missing
• inappropriate
• unexpected
• odd

Checklist questions

• Given positive/negative items on a checklist, only 3 patterns of responses are possible
• -all positive (positive person)
• -all negative (negative person)
• -both positive and negative (mixed feelings)

Subgroup Analysis

• analyzing subgroups provides deeper insights
• can take to any level of detail
• think about the data from the research end user's perspective
• look for the unexpected
• be open-minded
• be sensitive to sample size

Data Analysis

• demographic information:
• age
• gender
• purchase intent
• commercial believability
• commercial uniqueness
• reaction to commercial (two checklist items)

Slide 10