|
Mixed mode research is
possibly more talked about than done. And while a dozen or more software
packages now offer support for the method, users often find the hoped for
boost in efficiency still elude them due to irksome
technicalities.
There is no doubt that having their
own CATI and web-interviewing laboratory in the same building has helped
NIPO Software to polish up its well-established CATI system into provide
an extremely pragmatic solution to mixed mode research. The relationship
is both advantage and disadvantage, as buyers can be deterred from dealing
with a firm that is part of the TNS group. But this software company has
successfully managed such a Chinese Wall for many years and strives to
provide the same software and level of service to all its customers, both
internal and external.
The NIPO Fieldwork System offers a
vast sprawling network of rugged, workaday modules, more power station
than Porsche in its design ethos. But beauty is in the eye of the
beholder: this system, true to its Dutch roots, is designed to make money
out of fieldwork. The spec for interviewer PCs, for instance, is lower
than the PCs most people are throwing away today. It will run happily on a
Pentium ‘zero’. Screens are simple and uncluttered, though the supervisor
capabilities are fully GUI, and are very impressive.
On
the downside, at least for some, is that there is no GUI interface for
survey authoring – it is syntax only. But this is mitigated by a very
productive script definition environment with on-screen assistants that
suggest the range of syntax options, and instant previewing of what the
survey will look like in CATI or as a web survey, with all the routing
logic turned on.
What makes this software into a
compelling mixed-mode solution is the attention to detail that has gone
into making mixed mode research little different from setting up an
interview in CATI. |
They have strengthened the
handover points, with simple syntax commands to switch an interview from
one mode to another. Though CATI to web is the commonest switch, CATI to
CAPI, web to CATI, plus other combinations are just as feasible, and
interviews can be done entirely in parallel too, from start to finish. A
‘hard switch’ transfers all interviews at a pre-defined point, which is
ideal for a CATI interview followed by a self-completion module online, or
even a CATI recruit and screener followed by a web interview.
One database, securely inside the Internet
firewall, handles all the live data, and a web server, on the other side,
provides the engine for running the web interviews, but can safely pull
just the data it needs and push back the completed interview through the
firewall. It means that interviews can pass instantly from CATI interview
mode to web interview mode, in real time.
NIPO has even pioneered the ‘warm
handover’. It noticed that interviews tend to get lost during the handover
- sometimes because the respondent is really politely refusing by agreeing
to continue online, but sometimes because the email address is incorrect,
or the respondent is not sure what to do. The warm handover means the
respondent can get the email link, and the interviewer checks that the
first screen loads properly before signing off. As it also acts as a
prompt to the respondent, completion rates can be much higher when using a
warm switch.
The program may not be as visually
‘modern’ as some, but the pragmatic extensions to handle mixed mode
interviewing definitely lifts this program into the next
decade.
|
THE USER VIEW
FIELDWORLD ON THE NIPO
FIELDWORK SYSTEM
FieldWorld, a
fieldwork company in Amersfoort, Netherlands, has had considerable
success applying The NIPO Fieldwork System mode switching
capabilities to a diverse range of CATI and Web surveys, both to
boost response rates and reduce cost.
“It gives you several
advantages over a one channel system,” explains FieldWorld’s MD Bert
Buitenkamp, “because you can use it in situations where you need to
interview consumers where panels are insufficient, such as groups
with a very low incidence, or where clients want to have random
samples. And for B2B studies, we use it for refusers at the point of
refusal, to offer them the interview over the Internet and then send
them the email link immediately.”
By introducing such a ‘soft
switch’ at the point where an interview with hard-to-get B2B
respondents would normally terminate after a refusal, FieldWorld has
found that around 15% of those that would otherwise be lost to the
survey do log on and complete the interview in their own time, on
the web. While, for consumer surveys where sample is more abundant,
soft switches at the start of the interview generally see the
majority opting for a web survey, and between a quarter and a third
actually doing so, with net cost saving of around 30-40% over an
equivalent CATI-only study, and still usually a higher response rate
than telephone alone.
Buitenkamp sees the cost benefit as being
something driven by clients, but really of lesser importance than
the sampling and response benefits: the cost advantages will erode
once the method becomes more widely practiced, as he believes it
will. “Why we find it interesting is because, in the long term, the
industry as a whole will be helped if it can improve response
rates.” And there aren’t that many other remedies on
offer.
| |