Database Reference
In-Depth Information
“Wow! How are you going to do that?” Hans asks incredulously.
“Ah, that's where the fun begins. We'll run stepwise multiple regression and
determine the best model to increase adoption.”
Hans hesitates. “I have no idea what you just said, but it sure sounds cool. When
can we have the results?”
“Give me a couple of days to crunch the data. How's next Tuesday?”
“Sounds good!” he says as he stands up. “One last thing though.”
“Yes?”
“This time I'm going to bring you into see Joey to present. When I presented your
PowerPoint last time, he kept asking me stats questions I couldn't answer. You up for it?”
Despite the sudden surge of adrenaline, you calmly reply, “Sure, I wouldn't mind
inally meeting him.”
“Oh, you'll meet him alright,” Hans chuckles as he throws his laptop over his
shoulder. “Will you ever meet him! Have a great weekend!”
10.3 MULTIPLE REGRESSION
In this section, we shall introduce a small data set example using multiple regression
and use it to explain various aspects of the techniques and interpretation of its output.
In this small example, we have a Y and three X's, and a sample size of n = 25. In this
data set, Y represents the time it takes to complete a Major Task (in seconds) and the
three X's represent the time it takes the respondent, respectively, to complete three
component tasks. The goal is to determine which skills/types of knowledge have
more bearing on the ability to complete the Major Task. Later in the chapter, we shall
apply multiple regression analysis to the Behemoth.com data set with its “Y” and 15
independent variables (X's) and sample size of n = 180.
10.3.1 EXCEL
The data set of tasks times in seconds is shown in Table 10.1 .
SIDEBAR: KEEP ' EM TOGETHER IN EXCEL
One important thing that needs to be mentioned is that in Excel, it is critical that all of the X's be
contiguous (i.e., in adjacent columns—columns right next to each other); this will not be necessary
in SPSS. Indeed, the irst three columns, A, B, and C, display the three X's, and they are, obviously,
contiguous—right next to each other.
As when we performed a simple regression, we open “Data Analysis” to “Regres-
sion,” getting the dialog box in Figure 10.1 .
Note that we inputted the Y (dependent variable) range as D1 to D26, accounting for
the label in the irst row. For the X (independent variables) range, we inputted A1:C26.
(By doing this, Excel knows that we have X's in columns A, B, and C. The input process
would not work if the three X columns were not next to each other [i.e., “contiguous”],
 
Search WWH ::




Custom Search