Impact of innovation platforms and information
sharing on nurturing of smaller innovation
platforms: A case study of Tanzania Dairy
Development Forum
Kennedy Macharia Kago
MSc research proposal to Egerton University, 19 March 2015

Presentation Outline
 Introduction
 Problem Statement
 Objectives
 Hypotheses
 Justification
 Scope and limitation
 Output
 Conceptual framework
 Methodology
 Study Area
 Sampling design
 Data collection
 Data Analysis
 Model
 Work Plan
 Budget

Introduction
 Efficient and effective functioning of the dairy industry
relies on good information flow (Smallholder Dairy Project,
2005).
 Need for information flow crucial in developing countries
with households reliant on dairy for livelihood.
 Poor information flow results in high transaction costs,
information asymmetry, and increased risks along the
dairy value chain (Van Rooyen & Homann, 2007).

 Innovation platforms introduced late 1990s to bridge the
gap between different value chain actors (Pali & Swaans,
2013).
 An innovation platform as an equitable, dynamic space
bringing heterogeneous actors together to exchange
knowledge and take action to solve a common problem
(Cadilhon, 2013).
 Potential of Tanzania dairy industry unexploited with over
70% of milk being produced by indigenous breeds
(Makoni et al., 2014).
Introduction

Introduction
(Omore, 2014) (Kago & Cadilhon, 2014)

Problem Statement
 Actors lack of awareness on where and how to access
information, (Poole & Lynch, 2003).
 Dairy actors have uncertainty on quality and reliability
of available information due to lack of mechanisms to
ascertain information accuracy (Kilelu et. al 2013).
 Linkages between actors are complex and weak to
transmit information and promote its accessibility (Tui
et al., 2013).

Objectives
General objective
 To understand how the structure of the Tanzania dairy
development forum influences information sharing
among dairy value chain actors and its impacts on
nurturing of regional innovation platforms in the
Tanzania dairy industry
Specific Objectives
 To evaluate the extent that the structure of the Tanzania
Dairy development forum influences information sharing
between dairy value chain actors within the Tanzania
dairy industry.

Specific Objectives…
 To examine the influence of information sharing within
Tanzania dairy development forum on nurturing of
regional innovation platforms in the Tanzania dairy
industry.
 To examine how the structure of Tanzania dairy
development forum influences nurturing of regional
innovation platforms in the Tanzania dairy industry.

Hypotheses
 H1: The structure of the Tanzania Dairy development
forum positively influences information sharing between
dairy value chain actors within the Tanzania dairy industry
 H2: Information sharing within and outside of the
Tanzania Dairy development forum positively contributes
to the nurturing of regional innovation platforms.
 H3: The structure of the Tanzania Dairy development
forum positively contributes to the nurturing of regional
innovation platforms.

Justification
 The study results will guide the nurturing of other
innovation platforms in Tanzania’s administrative
regions, and elsewhere
 Limited choice of monitoring and evaluation frameworks
makes this study very relevant to users, researchers, and
the Tanzanian dairy industry.
 The study provides a baseline view of the Dairy
Development Forum as a benchmark for other
evaluative studies on the DDF

Scope and limitation
Scope
 The study focuses on DDF as a national dairy
innovation platform and does not evaluate existing
regional dairy innovation platforms.
Limitation
 There is limited literature on nurturing innovation
platforms thereby limiting amount of information
available on this study.
 DDF has only held three meetings and its impact on
the Tanzania dairy industry might still be minimal

IP- Structure
• Membership
composition and
diversity
• Decision making
process
• Committees
• Source of funding
Individual Structure
• Type of chain
stakeholder
• Gender
• Level of education
• Indicator of wealth
External environment
• Legal and regulatory
framework
• Cultural norms
Conduct of IP members
• Information sharing
• Communication
• Coordination
• Joint Planning
• Trust
Value chain Performance
• Advocacy
• Capacity building
• Value Chain
Development
• Collective promotion
• Joint quality standards
• Research &
development
• Market information
• Arbitration of chain
conflict
• Limiting transaction
costs
• Setting concerted
marketing objectives
....Other objectives set
by IP
Structure Conduct Performance
Conceptual Framework
Source: (Cadilhon, 2013).

Study Area
 It is located in the East Africa
region, south of equator
 Agriculture is the major
economic activity depended
by 75 % of the population
and contributes 28 per cent
to the GDP.
 Lies above 200m above sea
level with 5,000M being the
highest altitude on top of
Mount Kilimanjaro

Sampling
 Listing of all participants in all DDF meetings
𝑆𝑎𝑚𝑝𝑙𝑒 𝑠𝑖𝑧𝑒 =
𝑧2 × 𝑝(1 − 𝑝)
𝑒2
1 +
𝑧2 × 𝑝(1 − 𝑝)
𝑒2 𝑁
N = population size (DDF participants)
e =margin of error / confidence interval (0.05)
z = z-score (at 95%confidence level, i.e. 1.96)
p = percentage picking a choice, expressed as decimal
(Assuming normal distribution, 0.5 is adequate) (Kotrlik &
Higgins, 2001).

Sampling ….
 Stratified random sampling for dairy value chain
actors.
𝑓 =
𝑛
𝑁
 𝑛
n = strata size
N = Population size
f = strata sample size
 Paired sampling will be used for non-DDF
respondents
Input Suppliers Producers Processors Research/
Academic
Development
Partners
Policy Makers

Data Collection
 Both qualitative and quantitative data will be collected.
 Tools: Focus Group Discussions, Key informant
Interviews and Individual Interviews.
 Likert Scales (58 Statements identified through literature
review, while some have been used in previous studies)

Likert Statements
1= strongly disagree, 2= disagree, 3=
undecided, 4= agree, 5= strongly agree
We are satisfied with the quality of
information we get from value chain partners
(Li & Lin, 2006)/
Previous study
The information we get from value chain
partners is reliable
(Martey et, al.
2014)
We use the information shared with us in our
activities
(Warner, 2007)
Information on the market is easily accessible
to value chain actors
Previous study

Data analysis
 ANOVA tests to evaluate the significance of mean
differences between the members and non-members.
 Chi square analysis to test the statistical significance of
relationships between categorical variables.
 Factor analysis will identify underlying variables to
explain the pattern of correlations within a set of
observed variables.
 SEM to characterize patterns between observed and
latent Structural, Conduct, and Performance variables
 SPSS v.22 and SPSS AMOS v.22 will be used

Model Specification: Structural Equation
Modelling (SEM)
Structure – Conduct
i. 𝐶𝑜𝑛𝑑𝑢𝑐𝑡 𝐹1 = 1 𝑅𝑒𝑔𝑖𝑜𝑛 + 2 𝐺𝑒𝑛𝑑𝑒𝑟 + 3 𝐴𝑔𝑒 + 4 + 5 𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛 +
6 𝐷𝐷𝐹 𝑀𝑒𝑚𝑏𝑒𝑟𝑠ℎ𝑖𝑝 + 7 𝑂𝑡ℎ𝑒𝑟 𝐺𝑟𝑜𝑢𝑝𝑠 + 8 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 + 9 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑡𝑦𝑝𝑒 +
10 𝐹𝑢𝑛𝑑𝑖𝑛𝑔 𝑠𝑜𝑢𝑟𝑐𝑒 + ℯ𝑖

Structural Equation Model (SEM)…
Conduct – Performance
i. 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝐹1 = 1 𝐶𝑜𝑛𝑑𝑢𝑐𝑡 𝐹1 + 2 𝐶𝑜𝑛𝑑𝑢𝑐𝑡 𝐹2 + 3 𝐶𝑜𝑛𝑑𝑢𝑐𝑡 + 𝐹3 + ℯ𝑖

Expected Output
 MSc thesis and award of a Master’s in Agricultural
Economics.
 A policy brief with result-based ideas usable by the DDF
secretariat to improve DDF performance.
 Publication of at least one research paper in peer-
reviewed journals.
 Generate results that will contribute towards the
refining and testing of a conceptual framework for the
monitoring and evaluating the impact of innovation
platforms

Research Work Plan
Month
Activity
May –
June
2014
August
2014
Septe
mber
2014
January –
February
2015
April
2015
May
2015
June
2015
Questionnaire Development,
Sampling
Data Collection
Proposal writing, Submission
Data cleaning, analysis, and
Interpretation
Thesis Compilation
Thesis presentation, corrections,
and Thesis Submission

Budget
Description US$/month
US$/6
months
Stipend (Includes accommodation, food and non-duty local transport) 1,000.00 6,000.00
Health Insurance 75.00 450.00
WIBA*(Group occupational Personal Accident cover) 20.00 120.00
IT 291.67 1,750.00
Security (student badge one off payment $6) 1.00 6.00
Space charges (Based on standard space/person 9SQM X unit cost $29.70) 534.60 3,207.60
Student travel (Economy return flight Nairobi to Dares Salaam) 83.33 500.00
Student travel Insurance (Estimate covers return trip above during 2days aftertravelling) 13.33 80.00
Field interpreter(TZS75000/day for16days - 10000forservice and 65000perdiem) 123.33 740.00
Local transport to otherregions (TZS20000/inter-city one-way trip for12trips +local taxis) 41.67 250.00
Overhead costs of local host institution (TZS500000/student forprinting, stationary, office space) 50.83 305.00
Communication costs in Tanzania (Internet, mobile phone, telephone) 33.33 200.00
Research Costs (SPSS software licence) 30.00 180.00
Supervision travel (Economy return Nairobi to Dares Salaam) 83.33 500.00
Supervision trip othercosts (Tanzania visa, perdiem and accommodation 1week's travel) 133.33 800.00
Sub total 2,514.77 15,088.60
Contingency (10%) 251.48 1,508.86
Total Costs 2,766.24 16,597.46

Acknowledgement of funding
This work was undertaken as part of the CGIAR Research Program on Policies,
Institutions, and Markets (PIM) led by the International Food Policy Research
Institute (IFPRI). Funding support for this study was provided by the CGIAR
Research Program on Policies, Institutions, and Markets and hosted by the CGIAR
Research Program on Livestock and Fish. This presentation has not gone through
IFPRI’s standard peer-review procedure. The opinions expressed here belong to
the authors, and do not necessarily reflect those of PIM, IFPRI, or CGIAR.

The presentation has a Creative Commons licence. You are free to re-use or distribute this work, provided credit is given to ILRI.
better lives through livestock
ilri.org

References
Cadilhon, J. et al., 2013. Innovation platforms to shape national policy. Innovation Platforms
Practice Brief 2., Nairobi, Kenya: ILRI.
Kago, K. & Cadilhon, J., 2014. Influence of innovation platforms on information sharing,
market access and nurturing of smaller innovation platforms: A case study of the Tanzania
Dairy Development Forum.
Klerkx, L., van Mierlo, B. & Leeuwis, C., 2012. Evolution of systems approaches to
agricultural innovation: concepts, analysis and interventions. In I. Darnhofer, D. Gibbon, &
B. Dedieu, eds. Farming Systems Research into the 21st Century: The New Dynamic SE -
20. Springer Netherlands, pp. 457–483.
Kotrlik, J. & Higgins, C., 2001. Organizational research: Determining appropriate sample size
in survey research appropriate sample size in survey research. Information technology,
learning, and performance journal, 19(1), p.43.
Li, S. & Lin, B., 2006. Accessing information sharing and information quality in supply chain
management. Decision Support Systems, 42(3), pp.1641–1656.
Makoni, N. et al., 2014. White gold: Opportunities for dairy sector development collaboration
in East Africa, Centre for Development Innovation Wageningen UR.
Martey, E. et al., 2014. Factors influencing willingness to participate in multi-stakeholder
platform by smallholder farmers in Northern Ghana: implication for research and
development. Agricultural and Food Economics, 2(1), pp.1–15.

Omore, A., 2014. Creating a livestock sector with global competitor advantages in East
Africa. In Agribusiness East Africa Conference. Dar es Salaam.
Pali, P. & Swaans, K., 2013. Guidelines for innovation platforms: Facilitation, monitoring
and evaluation, Nairobi: ILRI.
Poole, N.D. & Lynch, K., 2003. Agricultural market knowledge: Systems for delivery of a
private and public good. The Journal of Agricultural Education and Extension, 9(3),
pp.117–126.
Van Rooyen, A. & Homann, S., 2007. Innovation platforms: A new approach for market
development and technology uptake in southern Africa. International Crops Research
Institute for the Semi-Arid Tropics (ICRISAT). Bulawayo, Zimbabwe: ICRISAT.
Smallholder Dairy Project, 2005. Improving access to knowledge and information in Kenya’s
smallholder dairy industry SDP Policy., Nairobi, Kenya: Smallholder Dairy (R&D)
Project.
Tui, S.H.-K. et al., 2013. What are innovation platforms? Innovation platforms practice brief
1. Innovation platforms practice brief 1, pp.1–7.
Warner, J., 2007. Multi-stakeholder Platforms for Integrated Water Management, Burlington:
Ashgate.
References…

Annexes

Factor Analysis
 The basic factor model can be derived as follows;
 Assuming a column vector 𝒚 of scores on p measured
variables for a random individual, the common factor
model can be represented as;
𝒚 = 𝒙 𝒄 + 𝑥 𝑢
Where;
 is a p  r matrix of loadings for p measured variables on r common factors
 is a p  p diagonal matrix of unique factor loadings
𝒙 𝒄 is a vector of scores on the r common factors for the random individual, and
𝑥 𝑢 is a vector of scores on the p unique factors for the random individual.
All variables are assumed to have zero means. (Drton, Sturmfels, & Sullivant, (2007) )

Structural Equation Model
The structural equation model is conveniently divided into two parts: the
measurement model, and the structural (latent variable) model. The structural
model, also called latent variable model is;
ŋ𝑖 =∝ŋ +𝑩ŋ𝑖 + +Гξ𝑖 + 𝜁𝑖‚
Where;
ŋ𝑖 is a vector of latent endogenous variables for unit i,
ŋ𝑖 is a vector of intercept terms for equations,
𝑩 is the matrix of coefficients giving the expected effects of the latent
endogenous variables (ŋ)on each other,
Г is the coefficient matrix giving the expected effects of the latent exogenous
variables (ξ) on the latent endogenous variables (ŋ), variables, and
ξ𝑖 is the vector of latent exogenous 𝜁𝑖‚ is the vector of disturbances.
The i subscript indexes the ith case in the sample.
There is an assumption that E(ζi)=0, and COV(ξi′, ζi) = 0

Structural Equation Model
The measurement model has two equations; 𝑦𝑖
𝑦𝑖 =∝ 𝑦 +𝛬 𝑦 𝜂𝑖 + 𝜀𝑖
𝑥𝑖 =∝ 𝑥 +𝛬 𝑥 𝜉𝑖 + 𝛿𝑖
Where,
𝑦𝑖 and 𝑥𝑖are vectors of the observed indicators of 𝜂𝑖 and 𝜉𝑖 respectively
∝ 𝑦and ∝ 𝑥 are intercept vectors
𝛬 𝑦 and 𝛬 𝑥 are matrices of factor loadings or regression coefficients giving the
impact on the latent 𝜂𝑖 and 𝜉𝑖on 𝑦𝑖 and 𝑥𝑖 respectively
𝑦𝑖 and 𝑥𝑖 𝑖
and 𝛿𝑖 are the unique factors of 𝑦𝑖 and 𝑥𝑖
There is an assumption that the unique factors have expected values of zero, and
have covariance matrices of Ʃ 𝜀𝜀 and Ʃ 𝛿𝛿 respectively and are uncorrelated with
each other and with 𝜉𝑖and 𝜁𝑖.

Structure – Performance Path Analysis
𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒
= 1 𝑅𝑒𝑔𝑖𝑜𝑛 + 2 𝐺𝑒𝑛𝑑𝑒𝑟 + 3 𝐴𝑔𝑒 + 4 + 5 𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛 + 6 𝐷𝐷𝐹 𝑀𝑒𝑚𝑏𝑒𝑟𝑠ℎ𝑖𝑝 + 7 𝑂𝑡ℎ𝑒𝑟 𝐺𝑟𝑜𝑢𝑝𝑠
+ 8 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 + 9 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑡𝑦𝑝𝑒 + 10 𝐹𝑢𝑛𝑑𝑖𝑛𝑔 𝑠𝑜𝑢𝑟𝑐𝑒 + ℯ𝑖