Mobile-Based Technology for Monitoring and Evaluation

Submitted by on Tue, 04/12/2016 - 14:30

This knowledge product will be a reference guide for using mobile-based technology, and its associated benefits of real-time data sharing and data analysis; thereby enabling organizations, donors and citizens to use Monitoring and Evaluation (M&E) data for better implementation and delivery of projects.

1. Quick Start
Is mobile-based monitoring and evaluation for you? How to get started?

This section provides brief pointers for researchers, project managers, and donors on the possibilities and applicability of mobile technology to their work; and how to get started with mobile-based monitoring and data collection tools.

2. Mobile Technology: Options and Opportunities
How is mobile technology better than paper survey? What are your choices?

This section explains the various components that make up a mobile-based data collection system; lists their features and functionalities; and provides casestudies that highlight the different ways in which mobile-systems can be applied in the field.

3. Implementing Mobile
Technology in Monitoring and Evaluation (M&E) How to roll out a mobile system? What are best practices, dos and don’ts?

This section addresses issues related to implementation–planning time-lines, estimating costs, training field staff, and ensuring data quality and security. Using an in-depth case study, it underscores common practical challenges you are likely to face and offers solutions to minimize these.

Download File
Pdf as plain
mobile-based technology
for monitoring & evaluation
A Reference Guide for
- Project Managers
- M & E Specialists
- Researchers
- Donors
Page 2 of 43
Mitesh Thakkar, Founder and Director, Fieldata.Org
John Floretta, Deputy Director, J-PAL South/CLEAR
Diva Dhar, Policy and Training Manager, J-PAL South Asia/CLEAR South Asia
Nikhil Wilmink, Policy and Training Associate, J-PAL South Asia/CLEAR South Asia
Sree Sen, Policy and Training Associate, J-PAL South Asia/CLEAR South Asia
Niall Keleher, Director of Research Methods and Training, Innovations for Poverty
Michelle McConnaughay, Survey Programmer/Computer Assisted Interviewing
Coordinator, Innovations, Innovations for Poverty Action
Lindsey Shaughnessy, Data Coordinator, Innovations for Poverty Action
Page 3 of 43
Quick Start ............................................................................................................... 5
Mobile Technology: Options and Opportunities ....................................................... 8
Paper or Mobile? .................................................................................................. 8
Technology Options ............................................................................................ 10
Case 1: Multiple Mobile Options ......................................................................... 15
New Data Types ...................................................................................................... 19
Geographic Data ................................................................................................ 19
Case 2: GIS Data as Project Intervention ............................................................. 20
Multimedia Data ..................................................................................................... 21
Case 3: Multimedia Data for Verification ............................................................ 21
Electronic Sensors ................................................................................................... 22
Beyond Data Collection: Management and Outreach............................................... 22
Selecting the Right Technology ................................................................................ 24
Implementing Mobile Technology in Monitoring and Evaluation (M&E) ................ 25
Selecting the Right Technology Service Provider ...................................................... 25
Estimating Costs...................................................................................................... 28
Planning Timelines .................................................................................................. 28
Ongoing Activities: Monitoring of Projects........................................................... 29
One-time Activity: Surveys and Research Studies................................................. 31
Training and Piloting Tips 32
Ensuring Data Quality .............................................................................................. 33
Ensuring Data Security ............................................................................................ 35
An In-depth Case Study ........................................................................................... 36
Common Implementation Challenges ...................................................................... 39
Hardware and Devices Issues .............................................................................. 39
Software Issues ................................................................................................... 39
Logistics Issues .................................................................................................... 40
Data Related Issues ............................................................................................. 40
Tips and Best Practices ........................................................................................ 40
Reference: Software and Service Providers ............................................................ 42
Page 4 of 43
This knowledge product will be a reference guide for using mobile-based technology,
and its associated benefits of real-time data sharing and data analysis; thereby
enabling organizations, donors and citizens to use Monitoring and Evaluation (M&E)
data for better implementation and delivery of projects.
The Guide is organized into three sections:
1. Quick Start
Is mobile-based monitoring and evaluation for you? How to get started?
This section provides brief pointers for researchers, project managers, and
donors on the possibilities and applicability of mobile technology to their work;
and how to get started with mobile-based monitoring and data collection tools.
2. Mobile Technology: Options and Opportunities
How is mobile technology better than paper survey? What are your choices?
This section explains the various components that make up a mobile-based data
collection system; lists their features and functionalities; and provides casestudies
that highlight the different ways in which mobile-systems can be applied
in the field.
3. Implementing Mobile Technology in Monitoring and Evaluation (M&E)
How to roll out a mobile system? What are best practices, dos and don’ts?
This section addresses issues related to implementation–planning time-lines,
estimating costs, training field staff, and ensuring data quality and security. Using
an in-depth case study, it underscores common practical challenges you are
likely to face and offers solutions to minimize these.
Page 5 of 43
I’m a Field Researcher. I want to carry out a household survey (it’s very long!). Can I
use mobile-based data collection tools for it? How?
Yes, you can use mobile-based data collection tools for field-research – even for
long and complicated household surveys.
A good first step is to ask some of the mobile-based data collection serviceproviders1
to demonstrate their software or system or both to you, preferably
using your survey instrument itself. Once you get to see one of these systems in
practice, you can decide to set up the system in-house – if you have internal
technical resources, or out-source it to a service provider.
Key considerations in selecting a software platform and service provider:
♦ Survey question types: Can the software or service provider or both
integrate the kind of questions you require, such as tabular family rosters,
pre-loading of data, skips, validations, complex multiple choice questions,
location (Graphic Information Systems – GIS), media (photos), and open
questions? How quickly can a paper-based survey be converted into a
mobile format? How long does it take to modify or update the mobile-based
survey instrument later?
♦ Control and Management: Can you manage the software or system
yourself, or does it require ongoing support from the service provider? Can
you update your survey instruments; control field staff access to the surveys
and data; access collected-data in real-time; through the system? If you
have to manage it yourself, what kind of technical skills will you require inhouse?

♦ Monitoring: Does the software or system allow you to monitor and track
your field staff in real-time while data is being collected?
♦ Data Security: What kinds of data security features are included within the
software? Do these features meet your requirements?
♦ Time-frame and Costs: Does the software platform and service-provider fit
your timelines and budget? What kind of mobile devices will you have to
purchase? You may have to make some trade-offs in terms of your featurerequirements
vs. timelines and costs.

See reference list of Service Providers at the end of this document.
Page 6 of 43
I’m a Project Manager at a mid-sized NGO. I want to monitor the activities of my field
staff and projects. How can I use a mobile-based monitoring system?
An important factor in deciding how to deploy a mobile-based monitoring system
is to determine whether your field staff members are full-time or part-time
employees or volunteers. This will determine how much control you will have on
the mobile-devices that your field staff will be using for collecting and monitoring
In general, for projects with large numbers of field staff who are informally
connected to the organization, you can opt for an SMS-based system. This system
will work with any mobile device, and your field staff can use their personal
handsets to send in project data. This way, you do not have to invest in mobile
devices. The downside to an SMS system is that, other than ‘the phone number
that sent the data’ and ‘sent date-time’, you cannot audit the veracity of the
data, i.e. was it collected at the project-site? Were the numbers reported
correct? etc. Also, if large numbers of monitoring SMS’s are being transacted per
field staff, the running costs of an SMS system can add up.
For regular, full-time field staff especially at the supervisory and monitoring level,
you can provide smart mobile devices2
that allow for Mobile Apps with ‘formbased
interfaces’ for data entry, location (GPS) tracking, and media (photos) and
bio-metric data capture. The data collected through these Mobile-App based
systems can provide strong audit controls because of the location and visual
evidence that is electronically captured.
As part of your mobile-based monitoring activities, you can also collect data and
feedback from your beneficiaries directly. To facilitate their participation you can
consider toll-free IVRS (voice) or SMS services.2
IVRS and SMS services tend to be
resource intensive to set up, and running costs are transactional based, i.e. per
minute or per SMS.
Key consideration in selecting a software platform and service provider:
♦ Customizability: How customizable is the software platform to your specific
organizational needs? In general, it is avoidable to build a fully customized
system, unless you intend to have an in-house technical team to manage it.

See section on Technology Options.
Page 7 of 43
I’m a Donor. I want my NGO beneficiary to better monitor and share their ‘impact’
data with me regularly. What can I do?
With mobile technology, your NGO beneficiaries should be able to provide you
field-level data from their projects in real time. Having access to ongoing data,
instead of one-time annual report, allows you to understand the real impact your
support is having at the field level.
Also, by requesting ‘anytime and anywhere’ access to project monitoring data,
you can introduce a higher degree of transparency and accountability, at each
level, within your beneficiary NGO.
In general, monitoring and evaluation are seen as ‘non-programmatic costs’,
therefore efforts to make monitoring activities more transparent and rigorous are
limited. However, it is possible for organizations to switch to mobile-based
monitoring and management systems within their existing budgets. A good first
step would be to get your beneficiary NGO to estimate its spending on
monitoring and management activities. They can then try to look for a software
platform and service provider who can offer a monitoring solution within that
In addition to mobile-based data collection for monitoring projects, your NGO
may also benefit from integrating it with new social-media tools to reach out and
garner support for its work. Therefore, over time, rather than being a cost center,
technology-based monitoring and mobilization will be more of an investment for
your organization to grow and become more effective.
There are a few open-source software platforms and service providers that your
NGO can try out without incurring any costs (or minimal costs). A reference list of
software platforms and service providers is available at the end of the document.
Page 8 of 43
How do you decide whether paper or mobile is a better format for your monitoring
and evaluation activities?
Take a call based on which of the following criteria are important to you:
(a) Easy to develop and field-test questionnaires upfront
♦ Provides maximum flexibility in
formatting the questionnaire
♦ Can be shared and developed in
collaboration with others
♦ Can be printed and tested
Mobile or Tablet
♦ Follow standard formats – to fit
mobile or tablet screens
♦ Requires using software tools that
make it difficult to collaborate
♦ Can be tested only after deploying
mobile-devices in the field
(b) Allows for unstructured, impromptu notes and qualitative data
♦ Writing on paper is easier
♦ Field staff can fill in notes later
on paper-forms
♦ Responses can be written in
local languages
Mobile or Tablet
♦ Typing through keypad or
keyboard is slow
♦ Local language inputs at time of
data collection are difficult
♦ Handwriting recognition software
is still in its infancy
(c) Manage large surveys across regions with large number of surveyors
♦ Logistics of printing and tracking
paper-questionnaires is tedious
♦ Real-time tracking of survey
work is difficult
♦ Survey questionnaires cannot be
changed once deployed without
significant re-printing cost
Mobile or Tablet
♦ Can be deployed remotely to
mobiles or tablets. Survey work
can be tracked in real time
♦ If necessary questionnaires
can be changed even with
surveyors in the field
Page 9 of 43
(d) Data quality is paramount
♦ Cannot control or limit logical
flow of questions on paper
♦ Can have issues in deciphering
hand-written selections and
♦ Requires data entry – another
source of error
♦ Requires manual scrutiny of
surveys to check for errors and
missing data
♦ Effective monitoring–audit and
tracking, of data can be a
laborious and complicated
Mobile or Tablet
♦ Set logical question flow–thereby
making non-applicable questions
hidden from surveyor
♦ Set validation checks for answers
entered, prompt enumerators if
answers do not match ‘pre-filled’
previous data
♦ Some data cleaning is already
completed due to these features
built into the software
♦ Real-time data checking, allows
for prompt review of data quality
and makes auditing and
respondent tracking procedures
more nimble
(e) Cost and Time
♦ No one-time ‘hardware’ cost
♦ Ongoing costs of printing,
transporting and storing paper
♦ Data-entry operations take
significant time and
resources–training, data-entry
operators, transliterating local
languages, ensuring quality
through double data entry,
and reconciliation through
hard copy checks.
♦ Longer time-frame before data
is available for analysis
Mobile or Tablet
♦ Initial one-time cost of mobile or
tablet devices
♦ Additional costs for maintenance
such as batteries and replacement
due to loss of devices
♦ Ongoing data-plan costs, and
service-provider costs
♦ Real-time access to data to
monitor quality and progress
♦ Environmentally friendly as
printing surveys is avoided
Page 10 of 43
(f) Ability to collect new data types: Location (GIS), Media (Photos, etc.)
♦ Requires additional hardware
devices such as GPS devices,
cameras, etc., for collecting
non-text data types
♦ Non-text data types difficult
to integrate
Mobile or Tablet
♦ Single device with other
multimedia such as GPS, audio,
and video tools
♦ Non-text data can be integrated
with text data in real time
♦ Real-time access to location,
photos, etc., provide can collect
text as well as tie
♦ verification
Mobile data collection requires four components (Figure 1).
1. Hardware devices: To enter data into.
Mobile devices can range from ‘low-end phones’ that can only be used for
phone-calls and sending SMS’s, to specialized devices such as point-of-sale (PoS)
Add-on devices: Mobile devices such as smart phones can also be linked to addon
devices such as bio-metric sensors, bar-code readers, NFC and RFID3
chips to
record data such as finger prints, inventory tags, Smart Cards, etc.
2. Data collection software: To control how data is entered into the device based
on programmed formats and rules.
Data collection software is mainly required for smart phones, tablets and
notebooks, and it tends to be specific to the type of hardware device like
Android phones, Windows Notebook, among others. In some cases the software
is built-in with the hardware, for e.g. PoS terminals, or is not required, like in
low-end phones for sending data through SMS or IVRS.
Data collection software can be: (a) custom built; (b) licensed; or (c) subscribed
to as a service platform.

NFC = Near Field Communication; RFID = Radio Frequency Identification.
Page 11 of 43
3. Data transmission: To transmit or transfer the field-level data to a remote
location or a single central computer.
Mobile networks allow data collected in the field to be transmitted through SMS,
voice, mobile-internet, etc. With certain devices, like PoS terminals, data is
transferred physically by hot-syncing cables.
4. Data aggregation and analysis: To receive, collate and analyze data.
This can be done remotely through SMS, mobile-internet gateways on webservers
with online databases, or through local hot-syncing on local computers
using spread-sheet, database or statistical software.
Figure 1 and tables 1 to 4 provide details about the various features and options
available within these components.
Figure 1 Different Components of a Mobile-based Monitoring and Evaluation System
Page 12 of 43
Table 1: Hardware Devices: Options and Features
Options Low-end Phones Feature Phones Smart Phones Tablets Notebooks PoS Terminals
Cost $15–$50 $50–$150 $150–$300 $200–$400 $300–$600 $150 +
Screen Small Grey Scale Small Colour Touch Screen Large Touch Screen Large Screen Small Grey Scale
Data entry Keypad Keypad Touch Keyboard Touch Keyboard Keyboard Keypad
Make Calls Yes Yes Yes Depends4 No No
Send SMS Yes Yes Yes Depends4 No Depends5
Mobile Internet No Yes Yes Depends4 No Depends6
Internet (Wi-Fi or Cable) No No Yes Yes Yes No
Apps and Data Collection Software No Depends5 Yes Yes Yes No
Track Location (GPS) No No Yes Depends6 No No
Capture Photos and Media No Depends5 Yes Yes Yes No
Connect to External Devices (Printers,
Finger-print Scanners, etc.) No No Yes Yes Yes Yes
Hot Sync Data to Other Devices No Depends5 Yes Yes Yes Yes
Battery Life Full Day Full Day Half Day 8-10 Hrs 4-5 Hrs Full Day
Device Models for Reference Nokia 1203 Nokia C2 Samsung Galaxy Y Google Nexus 7 Acer Aspire One Custom Devices
4 Requires a SIM card slot.
5 Memory and functionalities vary with feature phone models.
6 Some models do not have GPS-capabilities.
Page 13 of 43
Table 2: Data Collection Software: Ownership Options
Custom Built Licensed Subscribed
Ownership You own it and can
change it
You own it, but
can’t change it You rent it
Set Up Time Long Short Short
Set Up Cost High High Low
Ongoing Costs Medium Low Low
Customizability High None Low
Upgradability Low None High
Stability Low High High
Table 2.1: Data Collection Software: Features Check List
Device or
Operating System
Which kind of devices or operating systems (OP) will the
software work on?
Usability How are questions, formats and rules created in the software?
Does it require programming skills? Can it be created in-house,
or does it require ongoing technical assistance?
Question Types What types of data (questions) can be collected by the
software, for e.g. text, numeric, single or multiple choices, date,
time, photos, location, etc.?
Formatting and
What types of display formats are supported by the software
like tables and matrices, sections, single-page grouping, hint
text, colored fonts, video or audio content, etc.?
Logic Functions Does the software allow for logical, rule-based actions such as
repeating of questions, skip rules, answer limits and validations,
pre-loading of data, randomization, etc.?
Deployment and
Can data-entry formats be updated remotely on the devices?
How much data can be stored on the mobile devices? Can the
data be edited after entry? Does it create back-ups?
How does the software control user access? Can multiple users
access the same questionnaires, edit data, etc?
Security Is access to the software password protected? Is the data
Language Support Can local (non-Latin) fonts be supported by the software for
displaying questions, as well as data entry?
Page 14 of 43
Table 3: Data Transmission: Options and Features List
Call IVRS7 SMS USSD8 Mobile Internet Internet Hot-sync
Set Up Time None 2-3 weeks 1 week 1 week None None None
Set Up Cost None High: Record audio files Low: SMS Gateway Medium: USSD G’way None Low: Cable or modem None
Ongoing Cost
Per Min
Per Exchange
Data Plan
Data Plan None
Demand Data from
Device (Pull)
calls out
Outbound calls
Outbound SMS
reminders Not possible Not possible Not possible Not possible
Receive Data from
Device (Passive)
Calls In-bound Calls Inbound SMS Receive USSD Code Yes Yes Cable Connect
Push Data to
Calls Out-bound Calls Outbound SMS No No No No
Send Data to
Device on Request
Calls In-bound Calls Outbound SMS Receive USSD Code Yes Yes Yes
7 Interactive Voice Response System: Voice instructions are provided over a phone call, and data is ‘punched’ into the phone’s number-pad in response to the
voice instruction, using number-based menu options.
8 Unstructured Supplementary Service Data: This is a menu-based SMS relay. It has to be initiated by the person with the mobile phone using codes that
normally follow the format like *123x#. It is a service normally used by large organizations such as phone companies and banks because it is cost-effective only
when subscribed to in bulk.
Page 15 of 43
Table 4: Data Aggregation Options
Remote Local
Hardware Web-servers Desktops or Laptops
Set Up Time Immediate to a day Within a day
Set Up Cost None A local Computer
Ongoing Costs Server and hosting costs Local Maintenance
Data Security
High: Sites are configured for
username and password-based
Requires stringent access
protocols on the local
Data Access
and Sharing
Easy: Create additional user
Difficult: Share password for
encrypted data

Data transmission to web-servers from mobile devices requires subscription to
mobile internet or data plans. Depending on the data-collection software, constant
Internet access should not be required to store the data. The data can be stored on
the device itself and transmitted whenever connectivity is available.
With local storage of data on the mobile device, unauthorized access and loss of data
can be an issue. Some software packages encrypt the locally stored data, as well as
store recoverable back-up copies of the data on the device itself.
We Save Big (WSB), a micro-finance institution needs to monitor the weekly
meetings of its Self-Help-Groups (SHGs). The paper-based weekly register-form filled
by its loan officers is as follows:
SGH ID: 53 Mth:Dec Year:2012 Repayments (if absent, enter ‘A’)
Date: 6th

Member Name Week 1 Week 2 Week 3 Week 4
1 Anita 350 500
2 Bina 290 A
3 Cony A 560
4 Deena 600 400
Page 16 of 43
We Save Big has 150 groups of 20 women each, and 10 loan officers who are in
charge of 15 groups each. Over the next 2 years, WSB expects to grow ten-fold, and
will have 30,000 women members, 1,500 groups and 100 loan officers.
Which one of the following three mobile technology options should WSB
Option 1
Device Software Transmission Aggregation
Low-end Phones None SMS Remote Server
(a) At the meeting, each SHG member who is present is asked to SMS the amount
they are repaying that day to a central SMS phone number.
Example: The SMS sent by Anita will be: r 500
(r = repayment; 500 = amount)
♦ The message will route through an SMS gateway and the remote server will
receive the data as: mm/dd/yyyy 9876543 r 500 (where 9876543 is Anita’s
phone number).
♦ If the remote server already has the master data with all SHG IDs, their
members’ names, and the members’ phone numbers, the data can be
tabulated on the website as:
Date SHG ID Phone No. Member’s Name Amount
mm/dd/yyyy 53 9876543 Anita 500
This data table can be used for report generation and analysis.
(b) Instead of the SHG members sending the data, the Loan Officer (LO) is asked to
SMS the following data to the server: r 53 anita 500
This data can be tabulated on the website as:
Date LO Phone
mm/dd/yyyy 3474234 Aysha 53 Anita 500
Page 17 of 43
Option 2
Device Software Transmission Aggregation
Smart Phone Data-entry App Mobile Internet Remote Server
Each loan officer has a smart phone with a data-entry App. Software. The following
questionnaire is available for data-entry:
This data gets transmitted over GPRS/3G, and when received on the server, the data
is similarly tabulated, as mentioned earlier.
Date LO Name SHG ID Member’s Name Amount
mm/dd/yyyy Aysha 53 Anita 500
Option 3
Device Software Transmission Aggregation
PoS Terminal Integrated Hot-sync Local Computer
♦ Each loan officer has a PoS terminal with a Smart-card Reader, and all group
members have Smart Cards.
♦ When repayments are made, the cards are swiped in the PoS, the Loan Officer
enters the amount repaid, and a printed receipt is given to the SHG member.
♦ At the end of the week (or day), the Loan Officer hot-syncs the PoS terminal to a
local computer when s/he visits the branch office.
Page 18 of 43
Option 1 (a): SHG members send their repayment data by SMS during the meeting in the presence of the Loan Officer.
♦ Will work with any mobile handset
♦ No hardware cost to WSB
♦ Low set up time
♦ Immediate access to data
♦ No additional costs for scale-up
♦ SHG members have to be literate and numerate
♦ Dependent on members having mobiles and numbers not changing
♦ Have to reimburse members for SMS cost
♦ Have to train members to use SMS
♦ Possibility of SMS entry errors
♦ Quality control difficult: members can send SMS anytime
Option 1 (b): Loan Officers send in SMS for each SHG member making repayments during the meeting.
♦ Will work with any mobile handset
♦ WSB has control over devices and costs
♦ Low set up time
♦ Immediate access to data
♦ Strenuous for loan officers to enter SMS for each member
♦ Higher possibility of SMS entry errors (longer content)
♦ Additional reporting will require more SMS formats
♦ Quality control difficult–Loan officers can send SMS anytime
♦ Scale up will require new, low-cost handsets
Option 2: Loan Officers have smart phones with Mobile Apps for filling questionnaires.
♦ Easy data-input interface, with error-checks
♦ Low set up time and immediate data access
♦ WSB has control over devices and cost
♦ Collect location or photo data for verification
♦ Better data quality
♦ Handset costs, especially when scaled-up to 100 loan officers
♦ Ongoing cost of subscribing to mobile data plans (GPRS/3G)
♦ Being expensive, higher possibility of handsets being stolen
Option 3: Loan Officers have PoS Terminals with printers and Smart-card readers.
♦ Ease of use: swipe card, enter amount, print receipt
♦ Members receive a ‘physical’ confirmation
♦ Better data quality
♦ High cost of PoS Terminals and providing Smart Cards
♦ Logistics for distribution, replacement and control of Smart Cards
♦ Data not immediate, Terminals have to be hot-synced
♦ Tied-in to Terminal provider for support and maintenance
Page 19 of 43
Mobile data collection offers three new data types which can be very useful for
monitoring and evaluation. These are as follows:
1. Geographic data – locations, paths, and boundaries
2. Multimedia data – photos, audio recordings, videos, etc.
3. Electronic sensors – fingerprints scanners, health-sensors, Smart-card
readers, decibel-meters, etc.
With smart mobile devices such as phones and tablets, you can capture geographic
data consisting of latitude, longitude and altitude of a point, a path or a boundary.
With these three data types, you can abstract additional indicators using GIS9
software to enhance your analysis. Some of these GIS indicators are:
(a) Locations and verifications: Mapping the location, path, area or boundary to a
geographical region like location of surveyors at time of survey on a district level
map; boundaries of a farmer’s land; administrative boundaries of a district, and
so on.
(b) Prevalence and density: Presence of certain activities or entities within a
geographical boundary, for example number of pharmacies in a slum
neighborhood; forest-cover (density) per square kilometer.
(c) Areas: Physical space occupied by certain structures, land, or activities, like
average land area of schools in rural vs. urban areas; area of agricultural land
encroaching forest land
(d) Proximity and spread: Distance between locations, instances and activities, for
example distance of village schools from tribal communities; distribution of HIV+
cases in reference to national highways, etc.
(e) Terrain: Geographical attributes of locations and regions, such as average rainfall
in areas with high incidence of malaria; life expectancy in high-altitude regions,
(f) Networks: Identifying and quantifying connectivity, for instance the duration
and distance of time between the first reported cases of Ebola; measuring road
connectivity and density in terms of total length and crossings, and so on.

Geographic Information System.
Page 20 of 43
(g) Change and progression: Changes over time in (or between) a given location or
region like rural-urban migration during the year; progression of students from
primary school to university, vis-à-vis, location, etc.
Project description and objectives
A randomized evaluation study was conducted in Delhi, India to test whether
providing information to government officials and slum dwellers can lead to higher
accountability and improved service delivery.
Two interventions were assessed as part of this project:
1. The effect of voter information campaigns on voter turnout and electoral
2. The effect of providing information on spending and quality of public services of
elected officials responsive during election sensitive periods
Role of mobile-based data collection
For the second intervention, field-based audits were conducted of garbage and toilet
services in slums using smart phones. This audit consisted of the following GIS
(a) Location and verification of garbage collection points and public toilets in the
slum areas
(b) Proximity and spread of public toilets to households and impact of toilet
cleanliness on household health
(c) Area of a city councilors’ ward which is classified as a slum
(d) Prevalence and density of expenditure on public facilities like street lights in 1
kilometer radius from the city councilors’ residences.
Page 21 of 43
These GIS indicators were mapped along with geo-tagged photographs (as shown
above) and were presented to city councilors and community-based organizations.
Photographs, audio recordings and videos provide rich qualitative data for
monitoring and evaluation activities. Smart mobile devices using built-in components
like cameras and microphones can capture this data easily. Multimedia data can also
be used for audit and verification purposes.
Project description and objectives
The Community Assistants Initiative (TCAI) was launched in 2009 by the Ghana
Education Service to improve educational quality in government primary schools by
providing Community Assistants as support staff to schools through the
government’s National Youth Employment Program.
of slum
Location of
public facilities
Proximity of facilities
to households
Garbage Dump
Page 22 of 43
The Program had a randomized evaluation in-built study to measure the impact of
the Community Assistants on children’s attendance and learning outcomes. This
required longitudinal tracking of close to 25,000 students over 2 years.
Role of mobile-based data collection
The TCAI research team deployed 120 field enumerators with feature phones to
photograph and track the 25,000 children every year. During the first baseline, a
database with individual pictures of the students by school was collated in real-time
on a web-server. This list with photographs became an audit as well as tracking tool
for the project. It allowed field supervisors to visit the schools to identify children
who had been surveyed.
Mobile devices such as smart phones and tablets have built-in sensors such as
accelerometers, microphones (can be used for measuring decibel level), light sensors,
etc. These devices can be further enhanced by external add-ons such as fingerprint
scanners, card-readers, motion detectors, air-quality sensors, etc.
With these kinds of sensors, monitoring and evaluation activities need not be limited
to data collected through human intervention; rather lot of remote conditions and
events can be measured by sensors and collected via mobile devices to better
monitoring efforts.
Mobile-based systems for monitoring and evaluation are mainly associated with data
collection, with information flowing from the field to a central management or
evaluation team. However, the same technology can also be used to improve day-today
operations, provide feedback to staff, and reach out to beneficiaries and other
Therefore, at an organizational level, mobile-based systems can and should be
conceived not only as data collection tools, but also as management and
communication tools.
Two critical organizational benefits of a mobile-based managed system are:
1. Taking informed decisions in real-time
2. Providing feedback and exchanging information between stakeholders in real
Page 23 of 43
Living Goods is the Avon of Pro Poor Products. It operates networks of independent
entrepreneurs who make modest incomes going to door-to-door selling affordable
and effective solutions designed to improve the health, wealth and productivity of
the world’s poor.
Living Mobile, their in-house mobile management system, is a cornerstone of their
scalability and sustainability strategy. By integrating Living Mobile into their door-todoor
delivery system, Living Goods is building an end-to-end platform designed to
increase demand, improve access, and reduce costs for delivering products that save
and change lives.
A large component of Living Mobile is a two-way SMS communication system to
interact with the micro-entrepreneurs and client-households. The SMS system is
used for the following:
♦ Tracking health indicators and treatments-required by the client households
♦ Sending automated treatment reminders to client-households
♦ Informing field staff about new promotions, training schedules, and other
♦ Communicating educational materials related to Living Goods health products
For managerial and supervisory staff, Living Mobile consists of smart phones used by
the staff for reporting their field-work, and also for accessing data which is live
regarding their work. The smart phones are managed through an online portal with
reports and dashboards.
Page 24 of 43
How do you decide which is the right technology for your monitoring and evaluation
activities? In practical terms, your options are limited to the following five
combinations of the various technology components:
Device Software Transmission Aggregation
1. Low-end Phones  Not Required SMS  Local or Remote
2. Low-end Phones  Not Required USSD  Remote
3. Low-end Phones  Not Required IVRS/Call  Remote
4. Smart Phones/ 
Required GPRS/3G
Local or Remote
5. Notebooks/ 
PoS Terminals
Required Hot-sync  Local or Remote
You can evaluate and select from these five options based on the following criteria:
Table 5: Technology Selection Criteria Low-end Phones + SMS Low-end Phones + USSD
Low end
Phones +
IVRS + Call
Phones +
PoS + Hot
1 2 3 4 5
Monitoring ongoing projects and staff     
Carrying out large surveys  
Verification/audit of field activities  
Communication/outreach   
Provide real-time data to field staff    
Collect GIS/Sensor/Multimedia
Real-time data analysis    
One-off activities   
Repeated ongoing activities     
Varying ongoing activities  
Scale of Implementation
Large number of internal staff     
Mainly external respondents   
Small number of internal staff  
Low set-up/hardware costs  
Low-running costs   
Local/regional constraints
Low literacy/numeracy of field staff
Limited mobile network availability
Limited mobile Internet availability    
Require non-Latin script  
Limited electricity/recharge for device    
= Mandatory  = Optimal  = Possible
Page 25 of 43
Having covered how mobile technology can be used for monitoring and evaluation
(M&E) activities, and the varied technology options that are available to carry out
these activities, this section focuses on how to implement a mobile-based monitoring
and evaluation system. It consists of a series of checklists and dos and don’ts for the
various steps and processes entailed in the mobile-data collection process.
Given that mobile-based data collection consists of varied components, it requires
working with multiple technology vendors:
(a) Hardware providers selling phones, computers, etc.
(b) Software vendors for mobile applications for data collection and, if required, for
data collation and reporting
(c) Data-transmission service providers for SMS gateway services, IVRS-telephony,
and mobile-internet plans
(d) Web-hosting services for data aggregation and storage
In most cases, it will be easier and more cost effective to get a single service provider
to manage all these components for you, even if they in turn have to source some of
these services from other vendors.
With software being the core component, identifying the right software service
provider is critical – whether you custom build the software, license it, or subscribe
to it.
Key questions that you should ask of a potential software vendor:
1. Will their software work with the technology options applicable to you?
(a) Devices: Will it work with the specific models of the phones, tablets, or other
devices that you intend to use in the field?
(b) Data transmission: How will their software transmit the data from the
devices–SMS, USSD, GPRS or 3G, IVR-calls, hot-syncing?
(c) Data aggregation: How will data get aggregated? Where will the data be
stored? In what format will the aggregated data be available for analysis?
Page 26 of 43
2. Which kind of modules will their software contain?
Some generic modules that a mobile system is likely to contain are mentioned as
(a) Device application: Through which data will be entered and recorded.
(b) Form creation module: To create mobile forms and questionnaires that are
to be filled by field staff.
(c) User management module: To control who can fill the monitoring forms and
view field-level data.
(d) Data management module: For storing, viewing, exporting and importing
data from the devices as well as other sources.
(e) Report modules: To view data in generic formats such as map-based views
for tracking location data, etc.
(f) Security module: How secure is the software? What kind of security features
does it provide?
3. How easy is the mobile system to use?
Does the software require technical skills like programming? How are the
following activities managed within the software?
(a) Installation of the various software modules: Does the system work on a
website or does it need to be installed on local computers? How is
installation done on mobile devices?
(b) Creating questionnaires and forms: How much time does it take to create a
data collection form? Can multiple users collaborate in developing the forms
and questionnaires?
(c) Updating questionnaires: How often can surveys and forms be updated? Can
field staff access updated forms in real time?
(d) Managing and controlling access: How can multiple users access the
system? Can users be added, blocked, or removed easily? Is the system
secure and protected in any way? Can you monitor who is sending data in
Page 27 of 43
(e) Data aggregation: How is data from the field aggregated? Is it a manual or
automated process? What format is the data stored in? Can data from
multiple sources (not necessarily from the mobile devices) be integrated
together in the system? How?
(f) Data export and reports: What kind of formats can the mobile system export
the data in? Can data be exchanged with other software packages and
4. Will the device application support the data types and features you require?
Use your existing paper-based monitoring forms and surveys to identify the data
types and features you require. These can include:
(a) Question types: Text, decimal, integer, single choice, multiple choice, date,
time, location, photos, audio, video, barcodes, signatures, ranking, etc.
(b) Formatting and organization of question display: Grouping of questions into
sections, tables, or matrices; horizontal or vertical sequencing of questions
on a single screen; display of instructions or hints; displaying audio or video
content; formatted text–in bold, colors, etc.
(c) Logic functions: Does the software allow for logical, rule-based actions such
as–repeating of questions, skip rules, answer limits and validations, preloading
of data, randomization, compulsory answer required, etc.
(d) Language capabilities: Can the device application display and allow data
entry in local languages?
5. How does the software vendor price their services?
Will the software vendor develop a customized system for you? Will it be a
software package for which you buy licenses? Or will the software services be
provided through monthly subscriptions?
Irrespective of the type of engagement model you have with the software
vendor, you should clarify: One-time cost, recurring costs, cost of scalability
(from few to large number of users), maintenance and support costs, and costs
not included by the software vendor.
Page 28 of 43
A common and critical question in deciding whether to switch from a paper-based to
a mobile-based monitoring and evaluation system is how cost-effective is mobilebased
data collection?
Cost effectiveness of mobile-based data collection will vary from organization to
organization, and from project to project. However, the common costs to consider
while estimating the budget for a mobile-system are as follows:
1. Hardware costs: Are your field staff full-time or part-time or contractual
employees? How many field staff do you have? Is it better to purchase mobile
devices for your staff, or to incentivize them to use their own devices?
If you purchase mobile phones or tablets, you can amortize hardware costs by
assuming the devices usable lifetime to be minimum 2 years, and breakage or
loss rate to be around 5 per cent over the 2-year period.10

2. Data transmission costs: These are mainly recurring costs incurred by SMS, USSD
or mobile-internet (GPRS/3G) usage. In certain countries, you may also incur
one-time set-up cost for dedicated services such as SMS short-codes.
3. Data aggregation costs: These are recurrent costs that only apply if you use
remote, web-based data aggregation and hosting. Your technology vendor or
you will need to subscribe to a server-hosting (or cloud computing) service
4. Management costs: The mobile-system will require an internal or out-sourced
team to manage and support it. These costs will depend on the scale at which
the mobile-system is being implemented.
Planning timelines can be broken into two categories:
1. Ongoing activities like monitoring of projects
2. One-time activity like surveys

10 Based on feedback from projects undertaken by in 2012.
Page 29 of 43
♦ Planning and design phase
(a) The first step of any project should always be a thoroughly thought out
theory of change11 where the outputs and outcomes have indicators for dayto-day
operations and monitoring that are clearly identified and
(b) The format and frequency of how and when these indicators need to be
collected has to be mapped out. This usually involves a field monitoring plan
that details who will collect the data, when and how.
(c) Once this framework is in place, a mobile-based data collection plan can be
worked on.
(d) If you already have paper-based monitoring forms and formats ready, this
planning and design phase should not take very long. It is important,
however, to involve field staff in this phase so they both understand the
purpose and buy into collecting monitoring data.
♦ System set up
(a) This is a technical phase; you will have to work closely with your vendor to
set up the system as per your data collection requirements.

11 Refer to Chapter 4 of ‘The Road to Results: Designing and Conductive Effective Development
Evaluations’ by the World Bank Publication on How to Apply a Theory of Change to your M&E Activities, 01
June 2009.
Set Up
Phase Training Roll Out
Page 30 of 43
(b) Best practice usually involves setting up this system in incremental steps,
based on the feedback from the piloting phase outlined as under.
♦ Pilot phase
(a) During this phase, it is crucial to make sure the mobile system is accurately
capturing the data sent by field staff.
(b) Any kinks in the process should emerge during this phase and you should
provide extensive feedback to the technology vendor regarding the
timeliness, consistency and accuracy of the data coming in.
(c) Once the basic data-collection process has been piloted successfully and is
running smoothly, you can move on to training all your field staff on how to
use their mobile devices.
♦ Training
(a) Use the same mobile devices you will be using in your survey to train your
staff. Take your team through the complete process, including showing
them their data being received in real-time.
(b) It is useful to have a monitoring manual prepared as well as paper print outs
of the monitoring tools so that notes on how to collect monitoring data can
be recorded.
♦ Roll out
(a) In case you are providing the mobile devices to your field staff, you will have
to set up a logistics system to track your devices to prevent loss, damage or
(b) Furthermore, you may initially have to rely on alternate sources of data to
corroborate the timeliness and accuracy of the data being collected through
the mobile system. For example you might initially want to have field staff
noting down on paper what data is being sent each day, so that any
discrepancies between what is being sent and received can be highlighted.
(c) Discrepancies between sent and received data may occur. This could be due
to usability issues such as a human error; or some technical twists in the
system. As previously stated, thorough piloting and practice can help resolve
lot of these issues before data collection begins.
Page 31 of 43
♦ Developing monitoring reports
(a) Once your data-collection system stabilizes, you should have access to
regular, raw field data to manage your operations and assess your
(b) Automated reports from the data can be created either through your
software vendor or internally.
For surveys and research activities, timelines are likely to be driven by external
parameters, but you should plan for the following steps when estimating timelines:
1. Pilot your research instruments on paper
(a) Test your surveys and other data collection instruments as per your research
requirements first. This can be done using paper-based tools to ensure that
your questions are correctly framed and they elicit appropriate responses.
2. Set up and test the survey tools on mobile device
(a) Having finalized the content of your research instruments, programming
them on the mobile-device application is the next step.
Pilot survey
on paper
surveys on
Set up
devices and
Train and
surveys on
data from
Roll Out
Revise Surveys
Page 32 of 43
(b) Run mock surveys using a few devices to test question flow, connectivity
and transfer of data.
(c) Check the aggregated data to ensure that the received data sets are
3. Set up devices and test connectivity
(a) Install the mobile application on all the mobile devices that will be used by
the surveyors.
(b) If you intend to use mobile-internet for data transfer, ensure that the
phones can connect to the Internet. You may have to purchase SIM cards for
all the mobile devices.
4. Training, piloting and data checking
(a) If possible, the training should include piloting the survey instruments using
the mobile devices and software.
(b) This will familiarize the surveyors to the mobile device, the software and on
how to troubleshoot problems with both.
(c) Once the surveyors are comfortable with the device and the software, get
them to pilot the survey instruments and transmit the data in real time.
(d) Share with them the dummy data they have collected and elicit feedback
and questions regarding the data collection process.
(e) If your software allows it, revise your mobile survey instruments on-the-fly,
so that by the end of the training, the surveyors can take the mobile devices
with them containing the final version of the survey instruments.
♦ Initially use paper surveys with surveyors for them to annotate and make
comments on questions.
♦ Once the survey team has familiarized themselves with the content and
phraseology of the questions, the mobile devices can be introduced.
♦ It is always useful to have surveyors continue to refer to the hard copies of the
surveys to keep in mind training tips and to refer to longer questions that may
have been truncated in the software to fit the display screen.
Page 33 of 43
♦ Prepare a training manual specifically on how to use the mobile devices and the
software (along with a manual on the questions).
♦ Before the survey team is ready to use the mobile devices make sure the devices
have the software with the most up-to-date version of the surveys in it, based on
alterations made during field piloting and software testing.
♦ While training is going on, changes in the survey instruments will be inevitable.
Ensure you have enough time set aside to incorporate these updates in the
♦ Have the survey team trained in how to delete and re-install the software. This
way if they face any hardware problems, they can try to troubleshoot the issues
in the field.
♦ Instruct the surveyors to keep a separate record of the surveys they have
completed, saved and sent on a daily basis. Supervisors and project managers
should tally actual data received with the daily logs maintained by the surveyors.
Another option is to create mobile-based tracking forms for field supervisors to
fill with the number of surveys reported to have been filled by the surveyors.
This can be matched with the data received.
Data quality is enhanced when you can control or automate the flow of questions
and data-entry process on a device. This makes data collection Apps on smart phones
or tablets, IVR systems and PoS systems (with digital sensors such as Smart-card
readers) more accurate than SMS entry on low-end phones.
To improve the data quality of SMS-based reporting, the following steps can be
♦ Limit the number of fields required to be entered and reported through SMS.
Breakdown large number of fields into multiple SMS messages, rather than
squeezing them all into a single SMS.
♦ Provide a paper-based SMS format sheet for field staff to refer to.
♦ Send out automated SMS responses from the server to inform the sender that
their data has been received and is correct, or that it does not conform to preset
parameters and should be sent again.
♦ Regularly track the errors that field staff make in their SMS formats or answervalues.
These errors can be highlighted during regular training sessions.
Page 34 of 43
♦ If possible, standardize the low-end mobile handsets used by your field staff.
Different handsets function differently and it becomes difficult to provide
common instructions that work across the various types.
Software Features
To improve the quality of data gathered through smart phones and tablets, the
data-entry software and questionnaires should be programmed with as many of the
following functionalities as possible:
♦ Sequential, single question display: This forces the surveyor to focus on and fill
in one question at a time.
♦ Skip/piping logic: Do not display questions that are not applicable. The
questions should be displayed based on dependencies of previous answers.
♦ Mandatory questions: These questions cannot be left blank or skipped.
♦ Input masks: Control the number and types of characters that are entered.
♦ Validation rules: Similar to input masks, these are programmable rules that
limit surveyors to entering ‘valid’ responses only, for example age limits.
♦ Pre-filled, pre-loaded or auto-complete lists: Specifically for identifier names
or codes, if the data to be filled is available in advance (e.g. list of school
names), it can be pre-loaded and displayed as a selection list, rather than have
an open-ended text entry question.
♦ Double-entry checks: Repeat critical questions and enable automatic matching
of the repeated answers to check for consistency.
♦ Answer confirmation: Prompts to confirm the answer that has been answered.
♦ Error feedback: If answers are incorrect, provide details of the error type.
♦ Post-completion review: Allow for review of answers after completion before
sending data to server.
With mobile data collection you will have access to data in real-time, so additional
quality checks can be made with the data while it is being received. Your analysis can
look for missing data and outliers in the data while your surveyors are still in the field
to take corrective action.
Page 35 of 43
Data security is important for the sanctity of your operations, your stakeholders and
your research participants. Your data may contain Personally Identifiable Information
(PII) (information that can be used to uniquely identify, or locate a single person) and
other sensitive data such as financial or medical information or both which should be
kept private and inaccessible to most users of the mobile data collection system.
The potential security lacunae in the mobile-based data collection process and the
possible corrective security protocols are as follows:
Potential Security Gaps Solution
♦ Data stored or backed up
on the mobile devices
♦ If possible, use mobile devices
that have ‘lock and unlock’
password protection
♦ For SMS data, ask field staff to
delete SMS from sent folder
♦ For smart phones, tablets or
notebooks encrypt locally
stored data
♦ Encrypt PII data at time of entry
♦ Readable data accessible
by intermediary systems
♦ Encrypt data transmission
♦ For SMS data, get SMS gateway
provider to sign confidentiality
♦ Unauthorized access to
data files
♦ Multiple copies being
made of the data
♦ Use secure web servers or local
♦ Maintain a single, common
repository that can only be
accessed by ID/PW
♦ Do not use common passwords
between users
♦ Track user-access to data and
actions performed like data
downloaded, etc.
Other general principles for data security:
♦ Differentiate PII and ‘consent forms’ from other information right from the
beginning when data is being collected.
Page 36 of 43
♦ Transmit and store PII separately. Use link or reference ids or both to match data
sets with PII.
♦ Designate limited individuals who will have access to PII, and have them sign
confidentiality agreements.
♦ If possible, avoid making hard copies of PII. Keep this data in digital format, at a
single location that is password accessible.
The Mother Literacy Project was implemented by Pratham, the largest NGO in India,
to explore how mothers' literacy and engagement with children at home can
influence their children's learning levels. The impact evaluation of this Project was
done by J-PAL12 South Asia. The evaluation was carried out, from July 2010 to August
2012, in 240 villages in the state of Bihar and another 240 in the state of Rajasthan,
covering a combined sample of 8,888 households.
1. Selecting the technology platform
After an initial paper-based census survey in both states a switch was made to a
mobile-based data collecting system, consisting of feature phones (Nokia 2730
handsets), GPRS-mobile Internet transmission, and remote data aggregation on a
web-based server. At that time, the decision to switch was experimental. It was
primarily driven by an early demonstration by a known vendor of the feasibility of
mobile-based data collection. After the completion of the study, the research team
concluded that the initial time and monetary cost of buying phones, and developing
and testing software was justified as survey operations ran much smoother without
having to deal with large sets of paper print outs, scrutiny and data entry of paperbased
surveys over the 1.5-year period of the study.
2. Estimating costs
The largest cost was the fee paid to the digital data collection company and their
team of software developers who designed the custom made software and remained
with the project throughout the survey periods as consultants. This was necessary for
trouble shooting any issues with the application on the phones and data server.
Additional expenditure included the initial hardware costs in terms of the purchasing

12 The Abdul Latif Jameel Poverty Action Lab.
Page 37 of 43
of 120 feature phones and extra accessories such as chargers, extension cords, etc.
For mobile-services, a contract was signed with a telecommunications company for
120 SIM cards, with monthly billing charged directly to the project. Since voice-calls
were not needed, incoming and outgoing calls were disabled, and only the mobileinternet
(GPRS) and text message facilities were activated on the mobile-plans.
3. Planning timelines
Three months prior to rolling out the baseline surveys, phones and mobile-plans
were purchased, software requirements were ironed out and tested, and the survey
questionnaires were piloted at the time of training. The end-line, conducted a year
and a half after the baseline, mainly required updates to the survey questionnaires
and piloting those updates as the research team was already familiar with the
software and phones being used.
4. Implementation challenges
The Project specific implementation challenges related primarily to conducting
surveys in rural areas, as well as software and handset-hardware issues.
Implementing the survey in rural areas of Bihar and Rajasthan meant that the
research teams had issues with phone connectivity. A rule was therefore established
that the surveyors would save the data on the phones while in the villages, and the
crucial step of sending and transmitting the data would be carried out by the team
leaders (not the surveyors) at the end of the day from the field office. The team
leaders also used paper-based tracking sheets to monitor saved forms, sent forms,
and forms received on the server. This was required to address problems with data
reported missing, i.e. data collected and saved, but not received on the server.
Hardware issues encountered related to phone-memory limitations in the Javaenabled
feature phones. This arose due to ‘pre-filling’ and ‘pre-loading’ household
information on to the phones. This was data that could be ‘fetched’ from the server
onto the phones so that village names, village IDs, household IDs and respondent
names would appear as options on the screens of the phones and the surveyors
would simply have to select the correct respondent.
Initially, data of all villages and households was pre-loaded into the phone. During
testing, however, the large amount of this data when loaded onto the phones via the
online server caused memory issues, freezing the phones. The solution was to add in
survey ‘rounds’ whereby data was again divided into rounds of six villages which the
teams would visit based on the survey schedule. This resulted in a smaller subset of
the data being downloaded on a weekly basis.
This issue highlights the importance of thoroughly testing the software using dummy
data before the survey. In the Mother Literacy Project, in spite of piloting studies,
Page 38 of 43
some memory issues arose which caused the loss of some surveys when phones
froze. To prevent this issue from occurring in the endline survey, memory cards were
installed in all the phones and the research associates were able to copy and paste
backups on the memory card if the application ever crashed thereby ensuring data
loss was minimal.
Finally another innovation which proved invaluable during the endline survey was the
creation of specialized digital tracking formats which the team leaders (monitors)
would have on their phones. When a household was completed in the field, the team
leader would enter a digital tracking survey on their phones to track:
♦ How many surveys were either completed or not completed.
♦ The reasons surveys were incomplete.
♦ What quality check was performed by the team leader on the survey (spot
check or full survey accompaniment).
♦ Whether a revisit was necessary due to incomplete information.
Occasionally households would have to be dropped from the sample and the reason
the survey was dropped was also recorded in this tracking format. This tracking
format was then compared to the actual data received on the server and any missing
questionnaires were immediately identified and sent for a revisit.
5. Data quality
In the field the fundamental mistake we wanted to control was surveyors mistakenly
selecting the wrong household on the device and thereby entering and saving
information and respondent answers under an incorrect household ID. Checks were
therefore put in the place in the form of double selection (or entry) of household
names and IDs to make sure this was never occurred. Also, household roster printouts
detailing the village information and IDs, landmarks, family members and the
required questionnaires to be administered to each household were also handed out
to the surveyors for them to track the data on the phones to the paper roster.
Data quality significantly improved because of the survey data being available in realtime,
at the end of the day on the web-server. This allowed the project managers to
assess productivity of the surveyors as well as the quality of their data. Also, with the
tracking sheets and back-checks, inconsistencies could be identified and addressed
while the surveyors were in the field to take corrective action.
6. Data security
Other than the census data, which was done before investing in mobile phones, all
survey data was collected and stored in electronic format. The mobile application
was password protected, and the data collected on it was saved in raw string format.
Page 39 of 43
While this data was not encrypted, it could not be interpreted without access to the
server-based questionnaire-formats. Once the data was sent to the server, it was
automatically deleted from the phone’s memory. The data received on the server
was password protected. This data was downloaded in csv/xls format, and upon
downloading it was encrypted.
♦ Purchasing of large number of mobile-devices: Local electronics vendors tend
not to carry large inventories. So, if purchases are not planned in advance, you
may have to buy different device models.
♦ Battery and charging issues: Mobile devices when used regularly require daily
charging. In rural areas, with limited electricity, this can be a problem.
♦ Memory issues: Some mobile devices have limited memory which may result in
data loss. Devices with expandable memory like SD Cards can address some
memory issues.
♦ Set up issues: Mobile devices differ in the way Internet connections are
configured. At times, mobile-network providers have to be contacted to assist in
setting up mobile-Internet connections.
♦ Device depreciation such as battery problems, and broken and loss of equipment
can be challenges. Putting penalties into the contracts of surveyors and monitors
to make sure equipment is properly looked after is a common way to ensure
equipment is well-looked after.
♦ Upgrades and fresh installations: If the software is being customized or upgraded
during deployment, it can create confusion with field staff and result in multiple
versions of the software being used in the field.
♦ Security: Data stored on mobile devices can be accessed if not encrypted or
password protected.
Page 40 of 43
♦ Multi-media handling: Large amounts of multi-media data like photographs, etc.,
can delay transmission of data.
♦ Set up issues: Activating SIM cards and mobile-internet facilities on hardware
devices can take time, especially when mobile network providers require
identification proof for each SIM card being issued.
♦ Billing and top up issues: It is difficult to track mobile usage by field staff, and
staff using pre-paid mobile plans may complain about running out of mobile
credits. On post-paid plans field staff may run up the bills by surfing on the
♦ Training: Field staff not only need training on the software being used, but also
on how mobile devices work, and how to use them.
♦ Field staff switching SIM cards: Tracking field staff based on their phone numbers
can become problematic because staff tends to switch SIM cards regularly.
♦ Theft: Mobile devices, being expensive and small, are easy targets for theft.
♦ Data loss: While rare, this can occur due to mishandling of the software, deletion
of the software or due to data transmission issues.
♦ Data of SMS forms are likely to have more errors compared to mobile App forms.
In addition to the many tips and best practices that have already been mentioned in
the document, the following points are important:
♦ Set up your mobile transmission requirements such as SIM cards, and mobile
Internet well in advance. This often takes time especially when large numbers of
SIM cards are needed.
Page 41 of 43
♦ In general post-paid connections (paid monthly) are easier to manage, as prepaid
connections need to be regularly recharged with credit. Field staff will not
be able to send data if they do not have credit on their phones.
♦ Get field staff to sign off on a mobile usage policy, which explicitly mentions the
cost burden of the mobile connections that will be borne by the organization,
and those that will be borne by the staff. This will prevent misuse of the mobile
connectivity for non-essential activities.
♦ Pre-piloting of survey on different mobile phones handsets is absolutely crucial.
This includes:
(a) Testing to see if the digital data collection application runs smoothly on the
(b) Making sure the correct local language scripts (where required) run on the
(c) Testing the survey thoroughly in the field and checking for bugs on different
handsets is vital because making changes once it is on the server is much
more difficult (piloting your survey can take up to 1 month depending on the
length of your survey).
(d) Customizations of your survey, making sure skip patterns and answer codes
run smoothly.
(e) Testing the memory of the phones by storing lots of data on the handsets to
check the memory of the phones is important.
(f) Sending completed surveys with ‘dummy’ data and reviewing the digital
data on the server is crucial especially looking at the format of incoming
data and whether it can be improved.
(g) Finding out the battery life of the phones and buying the required extension
cords to charge them extra batteries if required is necessary.
♦ When signing a contract with a digital data collection company, include a clause
that says online data should be stored safely, password protected and
encrypted/anonymized if possible.
♦ Finally a more general point: Paper vs. Mobile is not an either-or choice. You can
mix both methods of data collection to suit your needs.
Page 42 of 43
Software or
Service Model
Support and
Technologies Supported
 Email/Paid Support Computer-based None Local Statistics Netherlands
 Email/Paid Support Android, Java Devices GPRS/3G/Hot- sync Remote/Local Dimagi Inc.
 Online/Email/Paid Support Android, Java Devices GPRS/3G/Hot- sync Remote/Local Datadyne
EPICollect Free Email Android, Apple GPRS/3G Remote/Local
EpiCollect @
Imperial College,

 Online/Email/Paid Support Android, Java, Generic Mobile Phones GPRS/3G/SMS Remote Arthify Inc.
Freedom Fone
 Online/Email/Phone Mobile IVRS Local Freedom Fone
 Online/Email/Paid Support Generic
Phones SMS Remote/Local FrontlineSMS
Kobo ToolBox
 Online/Email/Paid Support Android Devices GPRS/3G Remote Kobo Toolbox
Page 43 of 43
Software or
Service Model
Support and
Technologies Supported
Nokia Data
Gathering  Online/Email Java/Windows Phones GPRS/3G Remote Various
Open Data Kit   Varies Android GPRS/3G Remote Various
OpenXdata  Varies
Android, Java,
Generic Mobile
GPRS/3G Remote OpenXdata
Pendragon Forms  Online/Email Android Devices, iPhone, iPad GPRS/3G/Hot- sync Remote/Local Pendragon Software
PoiMapper  Online/Email/Paid support Android Devices GPRS/3G Remote Pajat Solutions
RapidSMS   Varies Generic Mobile Phones SMS Remote/Local
Arthify Inc.
Dimagi Inc.
Caktus Group
Surveybe  Online/Email/Paid Support Computer-based None Local Surveybe
Regional Center Tag
Resource Type Tag
Language Type Tag
Image Thumb