www.uis.unesco.org Innovation data collection: methodological procedures ECO - UIS Regional Workshop on Science, Technology and Innovation (STI) Indicators Tehran, Iran 8-10 December 2013 Luciana Marins, UIS
Jan 13, 2016
www.uis.unesco.org
Innovation data collection:methodological procedures
ECO - UIS Regional Workshop onScience, Technology and Innovation (STI) Indicators
Tehran, Iran8-10 December 2013
Luciana Marins, UIS
www.uis.unesco.org
Ch. 8 OM - Survey procedures
Guidelines - collection and analysis of innovation data;
Comparable results over time and across countries;
Particular circumstances may require other methodology comparability.
www.uis.unesco.org
The survey approach
The “subject” approach: • Innovative behaviour and activities of the firm as a
whole;
The “object” approach:• Specific innovations (“significant innovation” of
some kind, firm’s main innovation).
www.uis.unesco.org
Populations (1)
The target population:
• Innovation activities in the business enterprise sector (goods-producing and services industries);
• Minimum: all statistical units with at least ten employees;
• Classification by size:
• Small: 10-49;
• Medium: 50-249;
• Large: 250 and above.
www.uis.unesco.org
Populations (2)
Statistical unit:
1 country (ETH)
17 countries (AZE, BLR, CHN, COL, CRI, CUB,
DOM, ECU, LSO, PAN, PRY, PER, SRB, UGA, UKR, URY,
ZMB)
4 countries (IDN, MYS, PSE, PHL)
2 countries (HKG, TUN)
Enterprise group
Enterprise
Establishment
Kind of activity unit
Size cut-off point:
12 countries (BLZ, HKG, COL, CRI,
DOM, ECU, IDN, LSO, PRY,
PHL, SRB, TUN)
2 countries (PAN, PER)
6 countries (AZE, ETH, MYS, UGA, URY, ZMB)
1 country (CUB)
3 countries (CHN, PSE,
UKR)
Number of employees
Turnover
Number of employees /Turnover
Number of employees /Other
Number of employees /Turnover / Other
Source: 2012 UIS Innovation Metadata Collection
1
5
1
3
1
0
1
2
3
4
5
6
0 employee 1 employee 5 employees 10 employees 16 employees
www.uis.unesco.org
Populations (3)
The target population (cont.):
• Classification by main economic activity:
• (National industrial classification system);
• ISIC;
• NACE.
ISIC Rev. 4 INDUSTRY
01-03 Agriculture forestry and fishing
05-09 Mining and quarrying
10-33 Manufacturing
35-39Electricity, gas, steam, air conditioning and water supply; sewerage, waste management and remediation activities
41-43 Construction
45-82 Services of the business economy
84, 85Public administration and defence; compulsory social security and education
86-88 Human health and social work activities
90-93 Arts, entertainment and recreation
94-99Other service activities; activities of households as employers and of extraterritorial organisations and bodies
www.uis.unesco.org
Populations (4)
The frame population:• Units from which a survey sample or census is drawn;
• Based on the last year of the observation period for surveys;
• Ideal frame = up-to-date official business register NSOs;
• If the register forms the basis for several surveys (innovation, R&D, general business), the information can be restricted to innovation.
www.uis.unesco.org
Survey methods (1)
Mandatory surveys increase response rates;
Census or sample surveys?• Sample surveys - representative of target population
(industry, size, region) stratified sample;
• Census - costly but unavoidable in some cases.
www.uis.unesco.org
Survey methods (2)
Completion: Survey type:
Source: 2012 UIS Innovation Metadata Collection
7 countries (CRI, DOM IDN, LSO, MYS, PSE, PRY)
16 countries (AZE, BLR, CHN, HKG, COL, CUB, ETH, PAN, PER, PHL, SRB, TUN, UGA, UKR, URY,
ZMB)
Voluntary
Compulsory, notspecified
Compulsory,enforceable penalties
Compulsory, notenforceable
14 countries (HKG, CRI, CUB, DOM, ECU, ETH, IDN, MYS, PSE, PAN, PRY, PER, PHL, UGA)
4 countries (AZE, BLR, COL, LSO)
6 countries (CHN, SRB, TUN, UKR, URY, ZMB)
Sample
Census
Combined
www.uis.unesco.org
Survey methods (3)
Domains (sub-populations):• Subsets of the sampling strata;
• Potential sub-populations: industry groupings, size classes, regions, units that engage in R&D and innovation-active;
• Guidelines:» Same statistical units and classifications;» Consistence of the methods for results calculation;» Documentation of deviations in data treatment or differences in
the quality of the results from the domains.
www.uis.unesco.org
Survey methods (4)
Sampling techniques:• Stratified sample surveys: size and principal activity;
• Sampling fractions should not be the same for all strata;
Cross-sections: standard approach - new random sample for each innovation survey;
Panel data: alternative/supplementary approach.
www.uis.unesco.org
Survey methods (5)
Suitable respondents:• Methods: e.g., postal surveys, web-based
questionnaires, personal interviews;
• Unit’s most suitable respondent - very specialised questions that can be answered by only a few people;
• Try to identify respondents by name before data collection starts.
www.uis.unesco.org
Survey methods (6)
Data collection method:
Source: 2012 UIS Innovation Metadata Collection
8
2 2 2
1 1
2
1 1 1
2
0
3
6
9
In-person Mail Web In-person/phone
In-person/email
In-person/mail
In-person/web
Web/email/mail
Email/mail/
in-person/phone
Web/email/mail/
in-person/other
Other
www.uis.unesco.org
Survey methods (7)
The questionnaire:• Pre-test;• Simple and short;• Order of the questions;• Questions on qualitative indicators - binary or ordinal
scale;• International innovation surveys - attention to
translation and design;• Short-form questionnaires - units with little/no
innovation activity previously reported.
www.uis.unesco.org
Survey methods (8)
Combination of Innovation and R&D surveys: Reduction in the overall response burden;
Scope for analysing the relations between R&D and innovation activities;
Increase in the frequency of innovation surveys;
Country experiences - it is possible to obtain reliable results for R&D expenditures;
Longer questionnaire;
Units not familiar with the concepts of R&D and innovation may confuse them;
Different frames for the two surveys.
www.uis.unesco.org
Survey methods (9)
Survey combination:
Source: 2012 UIS Innovation Metadata Collection
6 countries (HKG, DOM, ETH, LSO, PAN, TUN)
1 country (CRI)
2 countries (PSE, ZMB)
15 countries (AZE, BLR, CHN, COL,
CUB, ECU, IDN, MYS, PRY, PER, PHL, SRB, UGA, UKR, URY)
Yes, R&D survey
Yes, R&D and ICT surveys
Yes, R&D and BusinesssurveysNo, stand-alone
www.uis.unesco.org
Survey methods (10)
Guidelines for conducting combined surveys:• Questionnaire: two distinct sections;
• Smaller individual sections;
• Comparison of results from combined and stand-alone surveys should be done with care - surveying methods should be reported;
• Samples extraction from a common business register.
www.uis.unesco.org
Estimation of results (1)
Weighting methods:• Weighting by the inverse of the sampling fractions
of the sampling units, corrected by the unit non-response;
• If a stratified sampling technique with different sampling fractions is used, weights should be calculated individually for each;
• Based on the number of enterprises in a stratum;
• International and other comparisons: same weighting method.
www.uis.unesco.org
Estimation of results (2)
Non-response:• Unit non-response: reporting unit does not reply at all;
• Item non-response: response rate to a specific question - % of blank or missing answers;
» Disregarding missing values + applying simple weighting procedures based on the responses received assumes that respondents and non-respondents are distributed in the same way biased results;
» Possibility: imputation methods.
www.uis.unesco.org
Estimation of results (3)
Unit non-response: Item non-response:
Source: 2012 UIS Innovation Metadata Collection
3 countries (CUB, ETH, SRB)
16 countries (AZE, BLR, CHN, COL, CRI, DOM, IDN, LSO, MYS, PAN, PRY, PER, UGA, UKR, URY, ZMB)
2 countries (HKG, TUN)
1 country (PHL)1 country (PSE)
Non-response survey
Re-contacting the firms
Re-contacting the firms andimputation
Re-contacting the firms andother
None
15 countries (AZE, BLR, CHN, COL, CRI, DOM, IDN, LSO, MYS, PAN, PRY, PER, UKR,
URY, ZMB)
2 countries (HKG, TUN)
1 country (UGA)
2 countries (CUB, SRB)
3 countries (ETH, PSE, PHL)
Non-response survey
Re-contacting the firms
Re-contacting the firms andimputation
Imputation
None
www.uis.unesco.org
Presentation of results
Descriptive analysis: no generalisation of results;
Inferential analysis: conclusions about target population;• Variance for the results: (average) values for innovation
indicators and their coefficients of variation and/or confidence intervals;
Results presentation: metadata (data collection procedure, sampling methods, procedures for dealing with non-response, quality indicators).
www.uis.unesco.org
Frequency of data collection
Every 2 years;
If not economically feasible frequency of 3 or 4 years;
Specify an observation period;
• The length of the observation period for innovation surveys should not exceed 3 years nor be less than 1 year.
www.uis.unesco.org
Annex A - 5. Methodological issues for developing country contexts (1)
Information system specificities:
• Relative weakness of statistical systems:» Absence of linkages between surveys and data sets;» Lack of official business registers;
• Involvement of NSOs;
• When lacking, basic variables about firms’ performance can be included in the innovation survey.
www.uis.unesco.org
Annex A - 5. Methodological issues for developing country contexts (2)
General methodological considerations:
• Survey application:» In-person;» Trained personnel;
• Questionnaire design: » Sections can be separated to allow different persons in the
firm to reply them;» Guidance/definitions;» Language and translation of technical terms.
www.uis.unesco.org
Annex A - 5. Methodological issues for developing country contexts (3)
General methodological considerations:
• Frequency:» Every 3 to 4 years (e.g., timed to CIS rounds);» Update a minimum set of variables every year;
• Purpose of surveys;
• Clear questions;
• Adequate legislative base;
• The results should be published and distributed widely.
www.uis.unesco.org
Basic innovation indicators:examples
www.uis.unesco.org
How do we measure innovation? (1)
Indicators - definition:
• Statistics and data, often gathered through specialised surveys, are the building blocks from which indicators are constructed;
• An indicator can be defined as something that helps us understand where we are, where we are going and how far we are from a specific goal. Therefore it can be a sign, a number, a graphic;
• An indicator quantifies and simplifies phenomena and helps us understand complex realities.
Source: International Institute for Sustainable Development / Adapted from Blakley, W. (2012). Providing and calculating innovation
indicators. Cape Town, South Africa. ASTII/HSRC/UIS Workshop.(PowerPoint Presentation)
www.uis.unesco.org
How do we measure innovation? (2)
Indicators - definition:
• Basic indicators: based on “one question”;
• Composite indicators: combine answers to several questions in order to examine a number of policy-relevant factors and better capture the diversity of innovative firms.
www.uis.unesco.org
Innovation indicators - examples (1)
Product or process innovation:
1. % of firms that implemented product innovation
2. % of firms that implemented process innovation
3. % of firms that implemented product or process innovation (innovative firms)
4. % of firms that developed in-house product or process innovation
5. % of firms that implemented new-to-market product innovation
www.uis.unesco.org
Innovation indicators - examples (2)
Product or process innovation:
1. % of firms that implemented product innovation
(N) =Number of firms that implemented product innovation (in each economic activity)
*100
(D) = Total number of firms (in each economic activity)
(N) =Number of Manufacturing firms that implemented product innovation
*100
(D) = Total number of Manufacturing firms
www.uis.unesco.org
Innovation indicators - examples (3)
Product or process innovation:
15
30
45
60
75
Brazil China Colombia Egypt Israel Malaysia Philippines RussianFederation
SouthAfrica
Uruguay
Product innovation Process innovation
Source: 2011 UIS Pilot Data Collection of Innovation Statistics
www.uis.unesco.org
Innovation indicators - examples (4)
Marketing or organisational innovation:
1. % of firms that implemented marketing innovation
2. % of firms that implemented organisational innovation
3. % of firms that implemented marketing or organisational innovation
www.uis.unesco.org
Innovation indicators - examples (5)
Marketing or organisational innovation:
3. % of firms that implemented marketing or organisational innovation
(N) =Number of firms that implemented marketing or organisational innovation (in each economic activity)
*100
(D) = Total number of firms (in each economic activity)
(N) =Number of firms in the Electrical machinery industry that implemented marketing or organisational innovation
*100
(D) = Total number of firms in the Electrical machinery industry
www.uis.unesco.org
Innovation indicators - examples (6)
Inputs:
1. Total expenditures on innovation (as a % of total turnover)
2. Expenditure on innovation by type of expenditure (as a % of total expenditure on innovation)
3. % of firms that performed R&D
4. % of firms that performed R&D on a continuous basis
www.uis.unesco.org
Innovation indicators - examples (7)
Inputs:
3. % of firms that performed R&D
(N) =Number of product or process innovation-active firms that performed R&D (in each economic activity)
*100
(D) =Total number of product or process innovation-active firms (in each economic activity)
(N) =Number of Services product or process innovation-active firms that performed R&D
*100
(D) =Total number of Services product or process innovation-active firms
www.uis.unesco.org
Innovation indicators - examples (8)
Key policy-relevant characteristics:
1. % of firms that were active on international markets
2. % of firms that co-operated with foreign partners on innovations
3. % of firms that co-operated with universities or other higher education institutions
4. % of firms that received public financial support for innovation
5. % of firms that applied for one or more patents
6. % of R&D-performing firms that co-operated with other institutions
www.uis.unesco.org
Innovation indicators - examples (9)
Key policy-relevant characteristics:
3. % of firms that co-operated with universities or other higher education institutions
(N) =Number of product or process innovation-active firms that co-operated with a specific partner
*100
(D) = Total number of product or process innovation-active firms
(N) =Number of product or process innovation-active firms that co-operated with universities or other higher education institutions
*100
(D) = Total number of product or process innovation-active firms
www.uis.unesco.org
Innovation indicators - examples (10)
Key policy-relevant characteristics: co-operation
Any type ofco-operation
partner
Other enterprises within your
enterprise group
Suppliers of equipment, materials,
components, or software
Clients or customers
Competitors or other
enterprises in your sector
Consultants, commercial
labs, or private R&D institutes
Universities or other higher education institutions
Government or public research
institutes
Brazil 9.7 1.1 5.0 3.5 1.0 1.9 1.9 n.a.
China n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Colombia 47.8 18.3 31.8 24.9 5.8 20.7 14.9 n.a.
Egypt 7.5 n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Ghana n.a. 28.1 21.1 31.6 17.5 22.8 12.3 8.8
Indonesia n.a. 37.8 66.3 n.a. 18.4 24.5 19.4 11.2
Israel 33.4 8.3 19.6 21.3 14.4 17.3 12.6 8.2
Malaysia n.a. 65.5 55.1 56.1 30.0 84.0 45.0 37.0
Philippines n.a. 91.2 92.6 94.1 67.6 64.7 47.1 50.0
Russian Federation 37.3 12.6 16.9 10.9 3.9 5.1 9.1 15.6
South Africa 33.0 14.2 30.3 31.7 18.6 21.1 16.2 16.2
Uruguay n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
EU-27 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Eurostat min 12.9 2.4 7.1 4.2 2.7 4.4 4.3 1.1
Eurostat max 56.2 23.0 41.5 36.0 30.8 33.9 30.8 26.3
Co-operation partner
Source: 2011 UIS Pilot Data Collection of Innovation Statistics
www.uis.unesco.org
Innovation indicators - examples (11)
Key policy-relevant characteristics:
3. % of R&D-performing firms that co-operated with other institutions
(N) =Number of R&D-performing firms that co-operated with other institutions
*100
(D) = Total number of R&D-performing firms
(N) =Number of R&D-performing firms that co-operated with other institutions
*100
(D) = Total number of R&D-performing firms
www.uis.unesco.org
Final remarks
Data collected with innovation surveys are a important component of comparative studies about countries’ competitive performance;
Strategically important for policy-makers;
Data confidentiality;
Data reliability.