close

Вход

Забыли?

вход по аккаунту

код для вставкиСкачать
SME Business Surveys
Andy Cosh
Centre for Business Research
Judge Business School
University of Cambridge
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
The central questions
• What are we seeking to achieve with an international
survey of entrepreneurship?
• If we simply want to benchmark across nations or
through time, then ‘filling in existing surveys may be
enough.
• Any effort to study individual firms over time, or to
explore policy impacts, is likely to need a more holistic
view of the survey needs.
• Any survey planning requires the answer to a set of,
often inter-related, questions.
• The following slides examine some of these questions.
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Key issues
•
•
•
•
•
•
•
•
•
•
Definition of entrepreneurship
Sampling frame
Sample size
Legal entity
Size range
Independents, divisions, subsidiaries
Sectors
Survey timing
Survey organisations
Survey method and cost
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Sampling frame
• Assuming businesses are tracked
• Two broad choices:
– National business registers
– Commercial databases e.g. Dun & Bradstreet
• Choice depends on:
–
–
–
–
Cost
Firm or establishment
Coverage
Timeliness
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Sample size
• CBR surveys have achieved responses of over 2,000
firms, but would like more, particularly if the survey is to
be longitudinal.
• One problem is the cell size when one performs size,
age, sector, legal form cuts of the data.
• Another problem is firm death (by failure for the poor
performers and by acquisition for the superior firms)
• Need a strategy for supplementing the panel.
• Need to decide whether to track entrepreneurs as they
leave the businesses – might be achieved by questions
to those remaining in the business (but what about
failures?).
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Legal forms etc.
• Can take a firm or establishment basis
• Self-employment/sole
proprietorship/partnerships
• Companies – restricted to private companies
only
• Independent firms only, or include subsidiaries
• Consolidated responses including subsidiaries
whether or not at same location
• Exclude not-for-profit organisations
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Size range
• Surely need to include micro businesses
• Probably need to include businesses with no
employees
• Upper size boundary, or constrain by age of
firm, or by whether a private business
• Need to recognise a growing trend towards
novel forms of business creation
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Sectors and stratification
• Should all sectors be included – doctors,
shopkeepers etc. can dominate the sample.
• Depending on the purpose of the study will need
to decide whether to over-sample in certain
cells.
• Similar questions arise concerning size ranges
and legal forms.
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Survey timing and organisation
• Want to be approximately simultaneous across the
participating countries.
• Expect to carry out over a three month period and avoid
key holidays in each country.
• Best to demonstrate official endorsement of the survey
whether or not the survey is actually carried out by a
government agency.
• Can be managed in-house, but there are many
specialised survey organisations available.
• Carry out non-response bias analysis.
• Re-survey every 2-3 years to retain survey teams.
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Survey method
• CBR has used face-to-face interviews only for pilot surveys and so I
have nothing to say about this method applied to major surveys.
• CBR has used both telephone and postal surveys.
• Telephone surveys have the advantage that they can be pursued
until they reach particular cell quotas and are efficient in that sense.
• Telephone surveys can include ‘real time’ checking of the
consistency of the answers.
• Postal surveys may be better for questions that require quantitative
answers – more likely to look up and less likely to guess.
• Response rates to telephone and postal surveys vary substantially
across countries and do depend on their perceived status and
relevance.
• For the CBR surveys the response rate has been broadly similar for
telephone and postal surveys.
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
Survey costs
• Some examples for a survey that took 30 minutes as a
telephone interview.
• Cost per completed interview:
– UK postal
– UK telephone
– US telephone
£20 - £25
£40 - £50
£85 - £100
• These costs exclude the cost of any rewards given.
• The CBR has never offered cash inducements but do
offer privileged access to the findings for respondents.
• The response rates are high when respondents to earlier
surveys are re-surveyed.
OECD Entrepreneurship Indicators Project Workshop
Paris Oct 2005
1/--страниц
Пожаловаться на содержимое документа