For Immediate Release

July 15, 2003

What amount of incentive will help raise my survey response rate?

Syosset, New York, July 15, 2003 - Beta Research Corporation, Erdos & Morgan Inc. and McPheters & Company have just completed a multi-phase test to continue our ongoing efforts to increase response rates for mail surveys.

What is currently used?
Incentives for mail studies have used amounts ranging from 25¢ to $1.00 to $2.00 to $5.00. Beta research has tested a variety of incentive amounts (see previous research) including, most recently, two $1 bills vs. a $2 bill.

Who signs the cover letter?
Often times, we overlook the importance of the signature on the letter or the stamp that is used. While we held the stamp constant (regular US 37¢ flag), we alternated two signatures across the studies:

Diane Clark, Director of Research
Beta Research Wave I, Erdos & Morgan Wave II

Nicholas Ferrari, Director of Research
Beta Research Wave II, Erdos & Morgan Wave I

What is being measured?
Our current test compares: 1) the impact of five $1 bills against a single $5 bill among a matched sample for 500 for a variety of questionnaire lengths (see chart below), and 2) a female non-ethnic signature vs. a male ethnic signature.

How is it being conducted?
Hoping to avoid "chance" differences, the same experimental design was set forth for Beta Research Corporation and Erdos and Morgan Inc. utilizing the same general questionnaire for each company.


PHASE ONE
Beta Research Corp.
Mailed 500 Per Group
 
2 PAGE
4 PAGE
8 Page
 
#
%
#
%
#
%
Five $1 bills
217
43.4
178
35.6
147
29.4
One $5 bill
206
41.2
161
32.2
128
25.6

PHASE ONE
Erdos & Morgan
Mailed 500 Per Group
 
2 PAGE
4 PAGE
8 PAGE
 
#
%
#
%
#
%
Five $1 bills
209
41.8
169
33.8
136
27.2
One $5 bill
204
40.8
149
29.8
120
24.0

A study of the demographic data revealed no significant differences.

PHASE TWO
Beta Research Corp.
Mailed 500 Per Group
 
2 PAGE
4 PAGE
8 Page
 
#
%
#
%
#
%
Five $1 bills
210
42.0
171
34.2
142
28.4
One $5 bill
213
42.6
152
30.4
124
24.8

PHASE TWO
Erdos & Morgan
Mailed 500 Per Group
 
2 PAGE
4 PAGE
8 PAGE
 
#
%
#
%
#
%
Five $1 bills
222
44.4
176
35.2
149
29.8
One $5 bill
211
42.2
159
31.8
128
25.6

What did we learn?
A) Five $1 dollar bills consistently generated higher response rates than a $5.00 bill. This result held for 11 of 12 paired comparisons.
B) A female non-ethnic name appears to help generate a higher response rate than an ethnic male name.



Beta Research
Syosset, New York

Contact:
Jason Gorelkin
Chief Operating Officer

P: 516.935.3800 Ext.248
F: 516.935.4092

jgorelkin@betaresearch.com