customers who looked for a job online and those who doesn’t ? INTERPRET · YOU DO THE SAME THING WITH REGRESSION BUT ALL · Like literally those who looked for a job online and those who doesn’t · Model validity + analysis · WEEK 11 MINI TASK
Cluster analysis / post hoc · Used to classify objects or cases into relatively homogeneous groups called clusters Produces segments that are homogenous within and heterogeneous between · Easily combine several variables to be the basis of segmentation · Segments might not be easy to identify with demographic variables · More advanced statistical techniques MAY BE required RQ: Is there a segment based on the variables ?____ Example Interpret: Size, profile, variable importance
· In terms of all variables, segment 1 has a lower mean value than segment 2. With the greatest difference in terms of being asked for information (segment 1 with 2.37 vs segment 2 with 5.50), being a good source of information (segment 1 with 2.31 vs segment 2 with 5.39), and lastly likelihood of taking a chance (segment 1 with 3.52 vs segment 2 with 5.24) · Thus, based on the mean comparison segment 2 can be named adventurers while segment 1 can be called followers · Since the highly adventurous segment compromises ~64% of HP’s customer base, this means that HP can produce highly innovative products Dummy variables · Represent variables as 0 and 1
· ONLY 2 GROUPS
You Can Do This OKKK Don’t worry about HD first
just do your BEST B
You've reached the end of your free preview.
Want to read all 44 pages?
- Two '16