Remote data collection in Papua New Guinea: an aid to policy deliberations

Mobile phone (Flickr/Ken Banks, kiwanja.net)

I often hear people say in Papua New Guinea (PNG) that it’s difficult to get accurate data. This comment is made in relation to all kinds of data: on banking, health services, health statistics, school enrolments, farmer activities, road quality, infrastructure expenditure and many more topics. It’s difficult to collect data due to rugged terrain, poor infrastructure and weak, costly transportation systems. The data that is available may have taken months to reach the relevant central government agency, may be incomplete or may be questionable in some respect.

Without accurate, reliable, timely data, it can be very tricky, therefore, for policy-makers to make decisions and plan activities. How can the best decisions be reached, without an evidence base?

Given the expansion of mobile phone network coverage and mobile phone use within PNG since competition was introduced to the sector in 2007, this creates potential for the use of technology to collect data remotely. There are two main methods that have been trialled and could be used to good effect in the PNG context: phone interviews and SMS data collection.

Phone interviews provide a number of benefits, versus the expensive, time-consuming, labour-intensive process of sending out field teams to conduct face-to-face interviews around the country. As has been demonstrated in a project in PNG (see paper by Kaski, Mursau and Maybanks), the telephone interview method can generate useful data from all of the provinces of the country in a relatively short space of time (just a couple of months, in this case). As has been found in other developing nations, there can be challenges with getting usable, working phone numbers and also with sampling (for example a possible urban bias, or a bias against people whose phone batteries may not have been re-charged due to financial constraints). Nonetheless, phone interviews can be an effective method of quickly gathering certain types of data.

SMS data collection has been trialled in PNG with two law and justice sector agencies (the trial was conducted by the Economic and Public Sector Program and funded by Australia). What the project showed is that SMS (short message service or mobile phone text messaging) can be a convenient, user-friendly way for people in disparate geographical locations to report data. Most district court clerks involved in the trial were happy to use the system, which involved answering a series of questions via SMS, rather than having to rely on fax machines or the postal system. An article on the project was recently published in the Commonwealth Governance Handbook (from page 34).

Forms installed on mobile phone handsets have provided another means for two PNG government agencies (National Department of Health and National Agriculture Quarantine and Inspection Authority) to collect data from field officers based across the nation. In both cases, the officers complete user-friendly fields on their phones, such as tick-boxes and drop-down menus, but the resulting data summary is sent via SMS. This system has the advantage of not requiring mobile internet, although a drawback is that if phones are lost, stolen or damaged, new phones with pre-installed forms need to be sent to the officers concerned.

So when people say to me that it’s difficult to get accurate, timely data in PNG, I find I’m frequently talking to them about the mobile phone-based data collection methods outlined here. Obviously these methods are not suitable for sensitive information – if you want to talk to someone about a traumatic experience they’ve been through, or if you want to discuss sensitive political issues such as the plight of West Papuans living in PNG, then it’s unlikely that remote data collection would be suitable for your needs.

But government agencies and non-government organisations often would like to have quick, regular updates on key pieces of information, such as stock supply levels or enrolment figures. So it seems to me that the methods discussed here could be of use, particularly when coupled with field visits, which would provide more in-depth analysis and serve other beneficial means as well.

I talked this week with managers of an international, non-government organisation which is running valuable programs in a number of rural areas across PNG. The ideas we discussed have coalesced in my mind in this way: perhaps they might do field visits every four months, say, to collect detailed information, take photos and build rapport with project participants. And in addition they could enhance their monitoring processes through conduct of weekly, fortnightly, or monthly remote data collection exercises, which would gather numerical data that requires regular updating. By synthesising the detailed data gathered through the two means (field visits and remote data collection), the organisation would have a much better understanding of the situation at each project site and the performance of its programs.

While some forms of information are best gathered through face-to-face discussions, visits to villages and field site inspections, other pieces of information can be collected remotely via phone interviews or SMS data collection. In the PNG context, such data collection efforts could speedily provide decision-makers with up-to-date information. Remote data collection could also prove to be much cheaper than other methods, most of which involve travel across the country.

Dr Amanda H A Watson is Mobile Communication Research Consultant with the PNG Economic and Public Sector Program. She is also a Visiting Fellow at The Australian National University within the School of International, Political and Strategic Studies. The views expressed in this post are the author’s own.

image_pdfDownload PDF

Amanda H A Watson

Amanda H A Watson is a researcher with the Department of Pacific Affairs, Australian National University. She taught at the University of Papua New Guinea’s School of Business and Public Policy under the ANU-UPNG Partnership. Dr Watson has also taught at Divine Word University, Macquarie University, Queensland University of Technology and TAFE NSW.

15 Comments

  • Nice share Amanda, you have covered the problems faced by people in data collection & also various solutions for it. The other means of collecting data in such area could be possible by using mobile forms that function seamlessly on tablets & smartphones, an app like Brew Survey uses offline mobile forms feature to collect data even when the device has no internet connection.

  • Thanks Amanda for sending me the link for this article. Very informative and sheds light into Mobile Phones as a key development aid towards collecting development related data for policy making.

    PNG is yet to appreciate the use of mobile phones as an aid towards development in terms of using it as means of data collection and information dissemination. Since the introduction of mobile networks in PNG, by far, SMS remains the most popular and cheapest form of modern communication. From automated data collection for one-off survey purposes to frequent SMS based reporting and provision of automated Q&A systems for frequently asked questions, SMS technology can be utilized in a variety of ways. The only major setback to rolling out development related SMS Services seems to be the cost per SMS for large volumes.

    Although this may not be appropriate, Matrix Consultancy is a local PNG SME that has invested considerable time and resources towards developing PNG MADE solutions using SMS. Mobile applications may become popular when cost of internet and internet enabled or application installable cost of mobile phones reduces but SMS will still remain the popular means of mobile phone based communication method for the foreseeable future in PNG. We have set up various SMS Services in partnership with PNG’s local MNOs. From complex and fully automated, live cloud based calculations involving various independent data sources and integration of popular sales/marketing and helpdesk systems down to simple google spreadsheet based SMS Services, we are now giving the opportunity for any PNG based client to be able to set up any SMS Service. You can find all updates of the latest SMS Services on our website.

    We had an interesting discussion last week with you and it was our pleasure in noting that we had similar development ideas using SMS technology. It was however disappointing to note that there is less awareness at senior bureaucratic levels in the government circles and even NGO/SME/corporate levels in terms of utilizing local resources. It may seem there is a general lack of understanding in the various options available for costing and features and these need to be explained and understood for such a technology to be appreciated and advocated as a key development aid.

  • Amanda, you have done well on using this method of data collection, it may work well for quantitative research methods. In quantitative research you are only collecting numerical data, and short answers that can be easily analysed. However, on the other hand, data as we know in PNG is not easily accessible because data is scattered and not easily stored in one particular place. For example, in the case of farm produce sold to supermarkets. The supermarkets receive supplies from farmers at random basis and from suppliers and not ant fixed amounts. that is an example of where researchers from outside PNG cannot understand and work out ways to analysis this kinds of data.
    Kind Regards
    Kuriya

  • Am greatful, am doing an assignment on importance of data in planning, your article is very suportive, will be one of my references. Its encouraging to have people with common interest in data collection. I suggest face to face interview for data sensitive issues and sms non sensitive issues, also perhaps 1-2 permanent local recorder/interviewer per village would help us on regular and updated data. moreover, its true collecting data in rural areas is a huge task. “A nation without records, history, data is a lost nation,”..by late Dr. Sam Kaima, PNG Achives, …..thanks B bange

  • Thank you to the ten people who have provided comments on this blog post.

    This week, I gave a presentation at the invitation of John Piel to advisers in the Provincial and Local Level Government Program (PLGP). While John and I had met before, it was through him reading this blog post that this opportunity came about. A number of the advisers saw potential applicability for strategic use of mobile phones in their work. I hope to help them to develop these ideas.

    Alistairs Dirua, I’d be happy to be in contact with you further about your work with health services and the use of SMS in collection of health data.

    Rod Reeve, I’ve been meaning to contact you to follow-up on your comment. I hope we can have further contact.

    Jason Brown, we used FrontlineSMS for the controlled trial of the SMS Story project in Papua New Guinea. While it has its limitations, FrontlineSMS is easy-to-use and can work well for sending of mobile phone text messages. There is also a new version, FrontlineCloud, which is operated via online management. For SMS surveys, there are a range of software and service provider options available.

    Thanks again to everyone who has read this post and particularly the ten people who’ve taken the trouble to post a comment.

    Amanda. 🙂

  • . . .

    I was going to recommend FrontlineSMS as an example of one company offering open source i.e. free ways of setting up mobile phone data collection.

    I see that Google is warning that it “may be hacked”, possibly due to what Sucuri identifies as an old wordpress installation. Sucuri and VirusTotal rank it as clean however.

    For the alerts, see VirusTotal here and Sucuri here.

    Meantime, here is an article about its background.

  • Thanks Amanda. Your innovation will work well for collecting government information (e.g. in your example of district court clerks), stock supply levels or enrolment figures. We do a lot of research in remote Australian communities to inform policy (as part of the Cooperative Research Centre for Remote Economic Participation (CRC REP)) and I wonder if some of our experiences can add to yours. Most of Australia’s remote communities have no mobile phone connectivity, so we use pre-loaded iPads (later downloaded) that are operated by local Aboriginal Community Researchers (ACRs). We have a network of around 200 trained ACRs – who are culturally empathetic and can work bilingually in an inclusive, respectful and genuinely consultative way. They are skilled at working independently or alongside communities and service providers to bring about positive impacts. They are mentored and trained by research leaders from UniSA, UNE, Curtin, CDU, Flinders and SCU. I’m sure PNG could mount a similar capability, even with a bit of help from DFAT (this may already be happening). It isn’t easy and it’s expensive on the surface, but it’s much more effective and cheaper than using FIFO consultants.

  • It’s true Dr. Amanda, I agree with you that mobile phone sms is one cheapest and fastest way of sending and receiving data on time. It truly covers the gap for poor internet services, poor postal services, poor telephone link etc…
    Practically, mobile phone sms is very helpful regarding timely data.

    Thank you

    Tina

  • I’m glad I stumbled onto this page and found people with common interests.

    I totally agree with Dr Watson. I represent the Christian Health Services in PNG. Most, if not all, of our health facilities are located in the remotest part of the country, and you can imagine the issues we are facing with Data Flow/Data Collection.

    For my case, SMS Data Collection is most definitely the way forward for us. The only problems I foresee is the User Training and the size/structure of data that can be allowed via text messaging.

    Time does not permit me to go into detail, but currently I am implementing DHIS2, a Health Management System, that has a module for SMS data collection.

    I just came here to say thank you for your article, and that it gave me the morale that I needed.

    Alistairs
    Christian Health Services PNG

  • An excellent write up and piece of work to keep people informed of the challenges in PNG related to data collection and of course how one can be able to manage these challenges to get the relevant data. The initiatives you have highlighted can become catalyst for relevant data for information analysis to inform better policy setting as you mentioned.

    A key challenge I believe is the ability of people with the data and the people requiring or demanding the data is for them to know that they can be able to utilize the methods you have discussed to better serve data collection in PNG for obvious purposes. This is probably a critical challenge and one that I personally think needs advocating across all public service sectors. Thus it will be good if we can integrate some of your discussions – especially generating data – into many of the other governance programs funded by development partners such as DFAT. I am sure these development (donor) partner programs will find it to their advantage to integrate some of the initiatives you have mentioned into their programs. You should talk to PLGP – a governance program targeting sub-national levels of government..

    Once again – thank you.

    John A Piel
    Finance Adviser

  • Thank you Dr. Amanda Watson for shining a light on our paper. Indeed, it is one of the most efficient methods if we have all contact data in place, all research participants are aware and there is excellent network coverage. If most data were collected from rural areas of PNG using this method to address Elementary Teacher Training Policy then I believe other sectors can also utilize this method.

  • Great summary of the potential for data collection using new technologies Amanda. Can vouch for the effectiveness in complex settings in PNG. With Amanda’s technical and professional support for the leadership of Executive Director of Village Courts & Land Mediation Secretariat in Dept of Justice & Attorney General, we collected data from 35 District Courts in many provinces in PNG for Village Courts related matters over 6 months. We had tried for 2 years with a hard copy system with 1 response. Mobile phone data project was invaluable to DJAG and demonstrated openness and capability of PNG agencies when supported well and practicalities of remoteness are addressed.

    Thanks Amanda.

  • Thank you Dr Amanda for this very resourceful information.

    The PNG government and policy makers need to take these methods into consideration because it it true that PNG needs concrete and real time data to make workable policies that will enhance equal distribution of wealth and services to citizens especially the majority in rural areas.

    Non-government organizations and international partners who are utilizing SMS technology to help people in the rural areas are helping and I commend them for that. Dr Amanda, your SMS projects in the justice, health and education sectors have proven greatly and this has to be promoted and supported by the PNG government and the departments concerned.

    A word of advice to individuals, non-government organizations and government agencies who do not know how to begin a SMS project to collect vital customer/client or population data/statistics. They ought to seek proper advice from experts like Dr Amanda before leaping into the mobile phone SMS technology for data collection. They could fall for a telecommunication company’s ‘marketing stunt’ but remember, it’s very costly and one might not get the data they needed.

    Thanks again Dr Amanda.

    …Regards…

    Stanley Mark

  • Very nice article. Collecting data in remote locations in PNG is not easy and it’s very expensive.

Leave a Comment