mvam: mobile technology for remote food security surveys...drc (katanga) drc (n. kivu) tajikistan....

mVAM: mobile technology for remote food security surveys

mVAM: technologie mobile pour effectuer des enquêtes à distancesur la sécurité alimentaire

September 16, 2014

http://www.wfp.org/videos/wfp-calling

What is the mVAM project testing?Que teste le projet mVAM ?

Live callsAppels téléphoniques

Text message

(SMS)

Interactive voice response

Serveur vocal interactif

Why are we piloting remote surveys?Pourquoi faire des enquêtes à distance ?

*Source: WFP article in ODI’s Humanitarian Exchange magazine, 2014

Performancecriteria

Face to face survey

Text message (WFP proof of concept, DR Congo *)

Live calls (WFP pilot in DR Congo and Somalia)

Surveyturnaround time (délaid’exécution)

6 weeks 1-2 weeks 2 weeks

Cost per questionnaire

$22 $3 - $5 $7 - $9

Security Exposesenumerators to risk

No ‘boots on the ground’, data collection possible in conflict areas

An affordable way to implement simple surveysUne façon abordable d’effectuer des enquêtes

face-to-face surveys remote surveys

DRC (Katanga)

DRC (N. Kivu)

TajikistanMauritania

Tanzania DRC (N. Kivu)

$-

$5

$10

$15

$20

$25

Face to face Voice Text message

Cost per questionnaire, North Kivu

Source: WFP data on actual costs from surveys conducted in 2013-2014

How does mVAM’s voice project work?Comment fonctionne le projet mVAM (partie vocale)?

Who are our partners?Qui sont nos partenaires ?

HumanitarianInnovation

FundInSTEDD

Information and research

firm

USAID GeoPoll Tulane University

Reduced CSI module (5 questions). Source: EFSA, mobile text survey

What we have learned: short SMS questionnaires and traditional surveys compare well – DR Congo, 2013

Nos résultats: les enquêtes par SMS concordent bien avec les résultats des enquêtes traditionnelles – RDC, 2013

What we have learned: voice surveys produce good quality, high frequency data – DR Congo 2014

Nos résultats : les enquêtes par téléphone donnent des résultats de bonne qualité et à fréquence régulière

Source: mVAM phone surveys, DRC

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Face-to-face Voice call Voice call Voice call Voice call Voice call Voice call

Nov-13 Feb-14 Mar-14 Apr-14 May-14 Jun-14 Jul-14

% o

f hou

seho

lds

Poor Borderline Acceptable

Source: WFP phone surveys

What we have learned: response rates to live calls,DR Congo and Somalia, 2014

Nos résultats : taux de réponse aux appels téléphoniques, RDC et Somalie, 2014

Source: WFP phone surveys, 2014

45%

68%

63%

55%

72%

73%

63%

58%

57%

0%

10%

20%

30%

40%

50%

60%

70%

80%

RO UND 1 RO UND 2 RO UND 3 RO UND 4 RO UND 5 RO UND 6

DR Congo Somalia

How do our response rates compare to other surveys?Comparaison de nos résultats avec d’autres méthodes

75%

68%

54%

73%

63%66%

52%

48%45%

57%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

Listening to Dar,Tanzania

Listening to SouthSudan

Listening to LAC,Peru

mVAM, DRC mVAM, Somalia

Source: World Bank, 2012; WFP 2014

(25 rounds) (4 rounds) (Wave 1) (6 rounds to date) (3 rounds to date)

What are our plans?Planification du projet

Phase 1, Voice: DR Congo, Somalia. HIF Grant.

Phase 2, Text message + voice: 10 ‘self-starter’ countries in different regions. USAID grant.

evaluation

Priorities for 2014-2015

o Evaluation of the voice pilot (Tulane University)o Mode test (research and information firm)o Ebola response pilot (Guinea, Liberia, Sierra Leone)o Continued compilation and dissemination of evidence

evaluation

June 2013 Dec 2014 mid-2015

mode experiment

mVAM and the Ebola food security responsemVAM et la sécurité alimentaire dans le cadre de

l’épidémie Ebola • Strong case for remote data collection, due to

restrictions to physical access and a need for higher frequency informationFort besoin d’enquêtes à distance, vu les restrictions au travail de terrain et le besoin d’informations régulières

• WFP will carry out 3 monthly rounds of remote data collection, from September to November, in each countryLe PAM va collecter des données dans chaque pays durant 3 mois, de septembre à novembre

• Technologies will first include IVR and SMS, then live calls

• Les méthodes par serveur vocal interactif et SMS serontd’abord utilisées, les appels téléphoniques ensuite

• Food security and market indicatorsSécurité alimentaire et indicateurs de marché

• Learning componentComposante d’apprentissage

Thank youMerci

• mVAM Video, “WFP calling: what did you eat today?”

• mVAM page on WFP site

• mVAM blog on Humanitarian Innovation Fund website

• Nesta guest blog, “Home-grown innovation at WFP”

• Article in the Guardian, “Using mobile phones to better understand refugees' food needs”

• CDAC Network guest blog, “WFP Calling! Lessons from Voice Surveys in the Democratic Republic of Congo and Somalia”

• Daily Development guest blog, “Press 1 if you have not eaten today”

• Humanitarian Practice Network article, “A new tool in the toolbox: using mobile text for food security surveys in a conflict setting”

• Article in the Africa Policy Journal (pp. 36-53), “Tracking vulnerability in real time: Mobile text for food security surveys in eastern Democratic Republic of Congo”

jean-martin.bauer@wfp.org@bauer_jm