A fortunate chain of events – a dry read

At Askia we love to talk about Askia things… and about a year ago, the technical team got together in a room and agreed on what was our biggest need: the ability to elegantly call a web service from a survey and decipher the result and store it appropriately.

Web-service not included

I have mentioned in previous articles how an API allows you extend your para-data. With the IP-address that you collect (and that we encrypt – GDPR is watching you), you can obtain the general location of the person. With the location, you can get the weather at the time of the interview and the likelihood they voted to a given party in the last elections.

You could always call a web service by adding some JavaScript in your page but that was not very elegant… and also made it hard to hide any authentication method.

So we decided to create a new routing where the Web Service was called from the server and not from the browser – effectively hiding the call from the interviewee. We got inspiration from the Postman interface and quickly put together a new routing.

The interface allows you to run different scripts depending on the success of the call and to manipulate and store the different parts of the response… and we introduced a new keyword CurrentHttpResponse.

QueryWebService

At that point, we thought that this had been relatively easy and we contemplated a well deserved visit to the local pub for refreshments.

XML and the Argonauts

As we were putting together an example – calling openweathermap.org to get the weather anywhere in the world – we hit our first problem.

The response looked like like this:

<?xml version="1.0" encoding="utf-8"?>
<current>
   <city id="6690581" name="Belsize Park">
      <coord lon="-0.18" lat="51.55"></coord>
      <country>GB</country>
      <sun rise="2018-03-06T06:33:58" set="2018-03-06T17:50:36"></sun>
   </city>
   <temperature value="282.33" min="281.15" max="283.15" unit="kelvin"></temperature>
   <humidity value="66" unit="%"></humidity>
   <pressure value="988" unit="hPa"></pressure>
   <wind>
      <speed value="2.1" name="Light breeze"></speed>
      <gusts></gusts>
      <direction value="200" code="SSW" name="South-southwest"></direction>
   </wind>
   <clouds value="40" name="scattered clouds"></clouds>
   <visibility value="10000"></visibility>
   <precipitation mode="no"></precipitation>
   <weather number="521" value="shower rain" icon="09d"></weather>
   <lastupdate value="2018-03-06T13:50:00"></lastupdate>
 </current>

To get the temperature, we would have had to look for the string “temperature value=” and extract the following digits… it was possible but a bit of a dirty hack, we felt. As stated before, at Askia we love to talk but we hate dirty hacks.

So we started talking about having a XML parser. The cool kids in the dev team took a clear stand: we do not need a XML parser and we would be a laughing stock if we implemented one. What we needed was a JSON parser. Even better we thought: what if AskiaScript could natively support JSON? Note: I can confirm it, we did a XML parser anyway – I hope you are not laughing.

JSON native and the dictionary

So we came up with the following syntax:

Dim myAuthorVar = @{
 “name”:”Jerome”,
 “age“:21,
 “occupation”:”laughing stock”,
 “busy”: true,
 “children”: [“Mackenzie”, “Austin”],
 “address” : {
    “postcode”:”SW12”
    “city”:”london”
    }
 }
Return myAuthorVar [“occupation”]

We were very excited but that meant we need a new variable type – it’s sometimes called an object or a map but also a Dictionary – the failed librarians and encyclopaedists that we are loved that… so there it was: the Dictionary. It allows to store a series of named values in one object. You can set its properties with a method Set like this myAuthorVar. Set (”Busy”, False ). And access them like you would with an array but by specifying a string instead of a number like this: myAuthorVar [“name”].

Variant and Arrays of Variant

I mentioned that it would be a good time to go to the pub when somebody asked what was the type returned by a dictionary accessor. In other words what was the type of myAuthorVar [“age”] ? The response to this is “it depends”… and there was no way of knowing before. Right now, it was a number, but if a web service had indicated “age” as “fifty-ish”, the result would be a string.

So we had to introduce a new type: the Variant

If you called myAuthorVar.TypeOf(), it would return “variant”… but inside the variant is a dictionary. So we created a method for Variant to know what was inside and we called it InnerTypeOf. myAuthorVar.InnerTypeOf() does return “dictionary”.

It was also nice to write @[1 ,2 ,3] or even @[3.14159,”pear”, “apple”]  – both are arrays of variants that we decided to call “arrays” for simplicity.

A variant could hold any of what we decided to call the base basic types: number, string, date, dictionary and any array of the types above. OK – let’s go to the pub! But then we remembered that JSON supported Null and Booleans… and because we wanted full compatibility, we had to create two new AskiaScript types: Null (which does not do much) and Boolean having the possibility of only taking two values: true or false.

Booleans and back compatibility

This was a can of worms – because we used to consider True and False to be numbers. Let’s imagine some script like this:

 Dim myVariable = (Q1_Name = 7)
 ' … some clever coding…
 myVariable = 42
 ' … more clever coding…
 If myVariable = 42 Then
    ' Save the world ...
 Endif

In classic AskiaScript, this would create a variable called myVariable as a number with a value of 1 or 0 and later taking the value 42 allowing the world to be saved.

We did not want to break back compatibility. I am going to summarize what was hours of discussions. We decided that comparators (like equal or Has) had to return numbers. If they returned booleans, the setting to 42 would now trigger an error because 42 is not a Boolean. And if we permitted an automatic conversion of numbers into Booleans, my Variable would take the value True (and not 42) which would change the way the scripts ran… and the world as we know would perish.

Wordy woes

Having spoken for so long, we were quite thirsty as you might guess. But we realised that our language would become very verbose and somehow inelegant if we had to convert Variants into the type we wanted whenever we wanted to use them.

In the example above, if we wanted to find out the length of our author’s post-code, we would have had to write:

 Dim hisAddress = myAuthorVar["address"]
 Dim hisAddressAsDic = hisAddress.ToDictionary()
 Dim hisPostcode = hisAddressAsDic ["postcode"]
 Dim hisPostcodeStr = hisPostcode.ToString()
 return hisPostcodeStr.Length

This was ridiculous… it would take ages to write any serious code… and we had better things to do than write verbose code (at that stage I was thinking of all the beers I would not be able to drink if I had to type that much to get my own postcode). So we went back to the drawing table and agreed that

myAuthorVar [“address”] [“postcode”]. Length was all we needed.

This elegant code was only possible if Variants supported ALL the properties and methods of ALL the basic types. That was a lot of unit tests that we had to do. So we focused (no blurred vision) and we wrote them.

This meant a serious rewrite and a careful management of conflicts: Format is a method for numbers and dates and they act very differently. So we put together a set of rules.

I’ll give you a reference

At that point, we had spent a lot of time on this, we were (very) thirsty but we wanted it to be perfect. And we realised we had a problem – what if we wanted to change the Postcode of our author (by code).

myAuthorVar [“address”] returned a Variant holding a dictionary with the address – a copy of the address. So to change the postcode we would have needed to write:


 Dim hisAddress = myAuthorVar["address"]
 hisAddress.Set ( "postcode" , "EC2A" )
 myAuthorVar.Set ("address", hisAddress)

This was again way too verbose. So we decided that accessors ( the closed brackets [ ] used by dictionary and arrays) would not return a copy of the address but a reference to the address of the author. This meant that we could write

 myAuthorVar["address"].Set ( "postcode" , "EC2A" )

This added very serious complication the the code  it’s called pointers as in dangling pointers in C++)… and that’s very difficult to make work. In the above example (as in life), the variable hisAddress can outlive myAuthorVar. We had to write a lot of unit tests to ensure that everything worked and we did not have memory leaks. This is discussed here.

For short, a variable stops being a reference as soon as you assign it something else.

AskiaScript Anonymous

We had an ongoing problem with the Value property of question – and we thought it’d be a good idea to address it now before we went to the pub.

Q1.Value returns a string if Q1 is an open-ended question. And it returns an array of numbers if Q1 is closed with multiple response. It can also be a number or a date…

Now let’s imagine we have a script like this

 Dim myVariable = Q1
 ‘ On Mondays at precisely 12 o’clock
 If Now. Day()= 1 and Now.Hour() = 12 Then
    myVariable = Q2
 Endif
 ‘ What is myVariable.Value here?

AskiaScript is a compiler – it wants to know the type of things before it’s run… but in that example, myVariable. Value could be of a different type depending on the day and time it was run.

And what if we had something like Q1.NextVisibleQuestion.Value?

So we decided that as soon as you put a question into a variable then the variable becomes an “anonymous question”. All methods of an anonymous question would work but the Value property would be a variant…. And we also decided to make sure that CurrentQuestion was an anonymous question. Problem solved! Drinks anyone?

But then we had another huge back-compatibility problem. Let’s look at the following code:

 
 Dim myVariable = QNumeric
 Return myVariable + 1

In classic AskiaScript, the system would add an invisible “.Value” after QNumeric (we call that an implicit property). myVariable would be a number and we would return that number incremented by 1.

But with the introduction of anonymous questions, myVariable was now a question. Facing an operator (the +) we would add again the implicit property .Value. But now value would be a variant and we had no rule to add a variant to something else… up to now.

So we made sure that we had rules to add any variant to another Variant – or any basic type or array of basic types. Not just add but also subtract, multiply, divide, compare – including all the keywords like  Has, HasNone etc. In total, combining 4 operators, a dozen of comparators, with 6 basic types and 3 types of arrays (number, string and variants) that made a lot of decisions to take (and a lot of discussions) and many many unit tests.

Before we started this development, we had 1667 unit tests ensuring that all functions of the AskiaScript behave the same from one version to another.

For this, we had to add 2231 (!) more unit tests. Once they all passed successfully, we added the whole thing to Suite 5.4.8 and we hope you’ll like it.

Enough Quant Tricks, we’ll be in the pub for a swift one – we deserved it.

Askia at Insight Show 2018

The Insight Show, one of the leading industry’s events in Europe, returns to London on 7 & 8 March and Askia will of course be there!

This year, in addition to our presence on stand IC24, we will be on the Insight Showcase stage to talk about Data Visualisation together with our partner and stand neighbor E-Tabs. It’s happening on Day 1, 7th March, at 14.40pm.

Registration is free for all Market Research professionals so there is no excuse to miss this opportunity to discover the latest industry trends, catch-up with your peers and stop by the Askia stand to talk about software in general and in particular: automation, APIs, community management and revolutionary dashboarding.

Interested? Then read the full program for more details and get in touch now to schedule an introductory chat or a demo with our team!

FACTS:

Insight Show, 7-8 March 2018, Olympia Central, London

Insight18_05854_Static Banners AW2

Come and meet us on Stand IC24

Presentation by Jérôme Sopoçko, Head of Development and Benjamin Rietti, CEO, E-Tabs on 7th March 14:40 – 15:00 on the Insight Showcase Stage: “Easy Visualisation of Market Research Data – The Quest for the Holy Grail”

Panel providers, unite – the speech at the ASC

On the 9th of November, the ASC invited some panel providers to attend a discussion on panel harmonisation. The discussion was orchestrated by Tim Macer.

Here was my speech – the written version at least as I may have ad-libbed a few unscripted things.

ASCPanel

Market Research is changing. You have heard it a million times – not in the way that Ray Pointer announced. There will be more surveys in 10 years than ever. That’s the good news. The bad news is that most of them won’t be run by MR institutes. The goose with the golden eggs is dead – client now run their own surveys which means MR companies – just to stay in business – have to be more competitive.

Goose with the golden eggs (before / after)

before-after

They started to delocalise in India, in Romania or in Ukraine. But that was not enough. To save more money, they have started to use automation.

This has its advantages – of course the surveys were a little bit more formatted… but Millward Brown had done that successfully for years. But once the bugs are eradicated, it’s efficient, fast and most of all cheap. And no blockade by disgruntled employees – although that’s more a French problem.

PresentationBrands

The problem is that end-clients are following up the trend – they can do automation too! They are using Zappi Store and Wizer… and SurveyMonkey and SurveyGizmo and ConfirmIt (and Askia). And ToLuna. And SSI self-serve. And Lucid. And Cint.

I have mentioned it at the ASC’s last conference: we have entered a golden age. The age of the API. A golden age for geeks like me at least: the internet is changing into a gigantic API where information is exchanged through web services. Everything is interconnected and uses the same interfaces.

IoT

I do not know if any of you have used IFTTT – If This Then That. It’s an app where you define a condition and an action. If I get near the house, put the lights on. If the temperature gets below 17 at night, put the heating on. If I enter the kitchen in the morning, put the radio on and start the coffee machine. If I have no milk in the fridge, order some. The IoT – the internet of things – is happening through one common interface through web services… and all industries are playing ball because they want their share of that big cake of a connected world.

oil-rig

I know we all have panel providers on stage so they might disagree with me. But panel data is no longer the only oil on planet Research. Customer databases are increasingly used because they can be energised by communities. And there is all sort of big data available at large – aggregated or not. It could be a loyalty card data, www foot prints or mobile phone data.

wine-glass

And just like for a good Bordeaux wine, to get quality you need to master the art of Blend. The merlot a bit dry and earthy – that will be your panel data. There is some cheap Merlot and very good Merlot too. And the Cabernet Sauvignon with its fruity flavours – that will be your behavioural data.

But unlike the IoT industry, Market Research providers have not decided to play ball. There are the ones who do not facilitate automation because they are afraid of losing control and burning panel. And there are the ones who do but work in isolation.

I do not believe there can be one company that will fill all the needs in Panel data. ToLuna is posturing itself as a one stop for all MR needs: the software, the panel and the behaviour. SSI is doing something similar and the merge with ResearchNow is going to be very interesting. The Leonard Murphy analysis about that on GreenBlog was great btw. And it won’t be scraps left for the others – because the need for data is growing – the need for specialised quality data will be growing too.

babel

But we need a common language. A common grammar. What is a social grade? How do I define national representativity? And how do I trigger a soft launch? How do I notify that a quota is full?

But there is another side to this discussion. If we let anyone access a survey which is tedious, long, repetitive, with grids, 2 max-diff exercises and one 20 minute trade-off, how do we reward the dedicated weirdos that filled that nightmare of a survey? How do we warn them that they are in for the long run? Because we might lose another goose with golden eggs. How can we stop the cull of panellists and the ever drop in response rates?

tediousness

I suggest we build metrics: number of questions, number of responses in a question. And then number of words per question, number of similar questions, number of mandatory open-ended questions… and then build a model.

$(Survey) => (Length(Survey) x TotalTediousness(Survey))-1

And then remunerate the panellists (and their providers) accordingly.

While I was preparing this discussion with all of you, most of you mentioned of how slow moving our industry was. It’s not just that: it’s protective, short-sighted and technologically unaware. And that’s everything the ASC is not. It’s at the ASC that triple-S, a format to exchange survey data between competing survey software was created and promoted. It’s two of my competitors, Steve Jenkins and Keith Hughes, who patiently showed my errors and taught me how to write a proper triple-S file. Let’s all be a little bit more like them and a little bit less like Apple who introduces a new plug and a new format with each new version.

chinese-propaganda

That’s my manifesto – a call for arms… please discuss and let’s move it forward.

MaxDiff grows!

This article provides an in-depth explanation of AskiaDesign‘s built-in capacity to manage MaxDiff data collection & analysis methodologies. For those of you who, like me, need a short reminder of what MaxDiff is; this is the definition provided by Wikipedia:

The MaxDiff is a long-established academic mathematical theory with very specific assumptions about how people make choices: it assumes that respondents evaluate all possible pairs of items within the displayed set and choose the pair that reflects the maximum difference in preference or importance. It may be thought of as a variation of the method of Paired Comparisons. Consider a set in which a respondent evaluates four items: A, B, C and D. If the respondent says that A is best and D is worst, these two responses inform us on five of six possible implied paired comparisons:

A > B,  A > C,  A > D, B > D, C > D

The only paired comparison that cannot be inferred is B vs. C. In a choice among five items, MaxDiff questioning informs on seven of ten implied paired comparisons.

MaxDiff table

We have recently added a new ADC to our offering that allows you to easily create MaxDiff tables in AskiaDesign. This article covers the setup process and usage for such comparison tables:

MaxDiff table ADC

 

This Askia Design Control allows you to easily create the required screen format for MaxDiff surveys. Add the ADC to your resources, drag it on to your Most response block, set any captions you want to appear in the headers of your grid and select the Least question it should be connected to. As with most ADCs, this survey control allows you to customise many parameters, such as:

  • Least Question: when you drag the ADC on to the response block for your ‘Most’ question, this is where you define which ‘Least’ question it relates to.
  • Most Caption: the caption you want to appear in the ‘Most’ column header.
  • Least Caption: the caption you want to appear in the ‘Least’ column header.
  • Centre Caption: the caption you want to appear in the centre column header e.g. this can be information about the loop iteration or screen number.

You can play around with this survey control in the following demos:

Alternatively, you can download (or even contribute) the MaxDiff ADC from Github!

MaxDiff interactive library

When conducting MaxDiff methodology you have a number of different parameters to consider and produce programming instructions for. At Askia, we have used the R software environment to do this for the different parameters and a large range of the options for each. We have created an interactive library in Design which asks you what option you want for each parameter. The result is a greatly simplified process for producing any MaxDiff design with Askia.

The available parameters are:

  • Number of questions: also known as the number of arrangements or number of screens. This is the number of screens the respondent will see during the course of the MaxDiff section.
  • Number of selectable items: this is the number of options to choose between per screen.
  • Number of items: this is the number of attributes or statements you want to include overall in the MaxDiff design.

As from version 5.4.6 of AskiaDesign, you can now use our Interactive Library feature to easily create and setup your MaxDiff design with the help of the above parameters:

MaxDiff interactive library

 

Check out the full article for more in-depth information & resources.

Adaptive MaxDiff

As we have seen in the above, the key point with standard MaxDiff is that the arrangements on screen are pre-set and do not adapt to the responses given in interview. In addition, the number of selectable options on screen is a constant.

However, in adaptive MaxDiff, the number of selectable options will change. Each round of screens, the items selected as Least are removed from the next round of screens. The number of items on screens therefore diminishes until you get to the start of the last round where you are asked to pick between all those you chose as Most.

The advantages of adaptive MaxDiff are that greater discrimination between items of importance is achieved. The disadvantages? Well, it could be argued that, since your initial answers create the upcoming arrangements, you do not have as much opportunity to change your mind about items you have rated least important in previous rounds.

This article details these differences, provides an example questionnaire to showcase the setup of this methodology with Askia as well as instructions on using and updating the example file for your own list of items.

New KB article roundup

This article aims to provide you with the best of our most recently published articles on our Help Centre, these range from AskiaDesign and AskiaSurf to AskiaWeb.

Redirect out of an Askia survey and back again

Sometimes it’s required to leave an Askia survey to take part in an external exercise and return to the survey to complete it. In such cases, it may be required to take parameters from the Askia survey to the external application or page. This article will show an example of these requirements using AskiaDesign.

Check out the full article for more details, access to the example survey and download all the attached resources.

Survey router

This article shows how to route a respondent from a main survey to two follow-up surveys out of a possible six depending on their initial selection and remaining SQL quotas. The seven surveys are set up such that the respondent will always be taken back to the correct position in any of their surveys if they close the browser and then click on the original link again.

The original article contains a link to a demo survey as well as an example questionnaire file in order to help you setup this methodology.

Quota logic examples in Design

This in-depth article provides a detail explanation of how to automatically manage quotas during fieldwork, specifically for complex quotas and/or for edge cases such as:

  • Sending an over-quota respondent to a short survey
  • Least Filled quotas

Quota logic example

Each case is fully detailed and provides example surveys to help you adapt the chosen method to your needs!

Local Storage

The Web Storage API provides mechanisms by which browsers can store key/value pairs, in a much more intuitive fashion than using cookies. This API provides two mechanisms:

  • Session Storage: maintains a separate storage area for each given origin that’s available for the duration of the page session (as long as the browser is open, including page reloads and restores)
  • Local Storage: does the same thing, but persists even when the browser is closed and reopened.

This article covers the use of localStorage as it is often used in CAPI surveys, where you want the agent to avoid re-entering the same data twice. A typical use case is an agent interviewing passengers on a single bus line. Once the agent has entered the bus line during the 1st interview, we want to pre-fill this question for new interviews, while leaving the possibility for the agent to edit at a later stage.

Check out the article for more details and access to the example questionnaires.

Capture browser’s user agent after every survey screen

The User Agent is basically an application that acts on behalf of a user. In the case of web browsers, it provides to the website / web application information concerning which browser, browser version, operating system, …

Askia only captures one instance of the browser’s UserAgent inside of the SQL database, meaning anytime you use the “Browser.UserAgent” keyword, it references the UserAgent that was captured in the database (which is the last device to enter into the survey). This Askia keyword does not keep track of which devices/UserAgents partook in the survey itself. Again, it only records the UserAgent of the last device that entered the survey or answered a question. If you want to keep track of which UserAgent was used to answer which question, you’ll need to use the snippet of JavaScript included in the article to pull in the UserAgent into an open-ended variable after every screen.

Improve speed of large Surf set-ups

This article sets out the steps needed for using askia Analyse & Surf to improve the (metadata) speed of Surf set-ups with a large number of .qes (wave) files.

We already had some more general tips to improve such rendering that would be useful for standalone datasets. However, these would not suffice in the case of complex AskiaSurf set-ups that comprise a large number of waves. The article therefore details the use of AskiaSurf’s Improve Metadata Speed feature.

Panel providers of the world, unite!

The short story

The industry is demanding more streamlining and automation… the only way that can happen is via standards – what are the Panel providers doing/proposing to do in this respect? We would like better visibility on their APIs and the differences between them… possibly talk about harmonising some key variables. We think there should be an automated standard evaluation of surveys in terms of length and complexity to better pre-evaluate the cost of sample.

We would like panel providers to explain their position – and their added values – in a (wait for it) panel discussion on Thursday the 9th of November in London – ORT House, London NW1 7NE as part of the one day ASC conference.

The very long story

I have always wanted to join an English gentlemen’s club. If I moved to the UK, I was going to be Phileas Fogg: travelling the word after a drunken boast and a wager over a bridge game. Last month (after 22 years in the country), it finally happened; I was asked to join the Association for Survey Computing.

I expected a standard acceptance ceremony: arriving blindfolded in a dark room, greeted by men in togas, a solemn oath with my hand on the 15th century preserved skull of the founder of the organisation, uttering something in Latin, maybe “Nam melius quaestiones”.

I was not disappointed. It was a Thursday morning Webex call to agree the subject of the November one-day conference. After the usual rambling about the weather (it was a cold September morning with a forecast for rain in the afternoon), roles were assigned. “You’re French”, they said, “you’re good at starting revolutions” they said “write a manifesto!”

And in truth, a revolution is needed. In previous years, the only way to have a lucrative MR business (not that I know about that) was to delocalise. The new trend is to automatise: you standardise a survey (want an ad test?), select the target (nat. rep. sir?) and you have your dashboard with your data ready just as your PayPal account is being debited. For this to happen, you need an automation platform (Zappi Store and GetWizer for instance) or a survey platform with an API… and you need a sample provider.

And that’s where it gets complicated.

A short digression into the real world

Let’s imagine you have built the perfect automated survey solution… it works nicely and you get results for every wave in exactly 2 hours 47 minutes. But for a given survey, you want to use a different panel provider to reach a very niche B2B target. You contact that specialist panel provider and explain your needs. They are enthusiastic about the idea and Adam, your contact there, wants to test your survey first – their panellists are special, you don’t get to burn their community like that. After 48 hours, Adam calls you back with a price, it’s on the expensive side but you agree right away because you want the data now – well you actually wanted it 45 hours and 13 minutes ago. Now he sends you a list of the internet parameters you need to accept in your survey… what was called SG with panel provider 1 is now called SocialGrade and GE becomes Gender3b… of course you already know why it’s called Gender3b; they introduced an “other” (and a “prefer not to say”) to the gender question. Your survey scripter says he needs a day (or two) to impact the changes… but he can only start after the week-end because it’s Friday and the web designer who did the icons for the gender question has already gone snowboarding for the week-end.

Here comes Monday, the designer damaged his knee and you decide to scrap the icons. The client checks the survey on Monday afternoon (they are based in the East Coast) and they want the gender icons back to verify the sample… so you add (early next morning) a nice routing to exit the survey if they say “other”. No soft launch, we don’t have time for that. Quickly (but not quite quick enough) you realise you have screened out 99% of respondents – your scripter wrote the routing the wrong way. You call a very unimpressed Adam to stop sending sample. Your guys finally correct the routing but unimpressed Adam has gone for the day. You eventually get through to him late morning the next day and he agrees to send more sample.

The data fills your automated portal nicely… you start to relax. You shouldn’t, your client has had a look at the data and he has noticed something very weird with the student segment. How is that possible? You’ve changed nothing there… until you decide to call Adam who reluctantly agrees to take you on. He explains calmly that although the internet parameter is indeed SocialGrade, the value 23 does not indicate “Students” but “Deep sea divers”… Did you not read the explanatory document he attached to his email on Thursday last week?

Now you know you are going to have an interesting conversation with both your client and your boss. But you may as well leave it until tomorrow.

And that’s how automation got scrapped in what you must now call your previous job.

The quest

So let’s get back to my personal quest – how can I make automation and surveys better? The answer is simple: by getting panel providers talking to each other.

That’s never going to be easy. Some of them are already panel aggregators and they feel they have already done the hard job. Others feel commoditising panels is not in their interest and will drive prices down. Some say it’s simply not possible because their own data is too rich. And all agree that sending sample to a broken or boring survey is the one reason that response rates – along with data quality – are dropping.

And they are right. Data is precious. We need to treat interviewees with respect and that’s not what we do when we send them a 40 minute conjoint survey (and tell them it will last 10). For panel providers to evaluate pricing properly, they need to know how good (and more likely how bad) our survey is.

We need to build metrics on the length of a survey (a lot of data is available there) but also on the boredom index of a survey: number of grids, number of responses per question, number of words per question text, number of questions with similar text, number of mandatory open-ended questions… and prices should vary accordingly.

Another option would be that the price could be fixed by the soft launch data. At the end of the survey, we measure interview interest and fix the price of the panel accordingly – with a rebate if the full survey data is actually below the early measure.

And how do we harmonise panel data? Should we break down questions in categories and sub-categories (demographics, lifestyle, political leaning) and incorporate that in the naming? Can we have the same break-down across different countries? For which questions? Should the naming convention clearly indicate the number of responses to avoid coding errors?

Be our panelist for a day

We’ve so many things to discuss… and we thought it’d be best if we did it in public. You, the panel providers, could tell us what you think… explain what’s special about your company, detail your API or your choice not to have one. And the ASC audience – rather technical but friendly – could tell you what they want and stand witness to your promises. The result could be a standard, (national or international), an API router or just an Excel spreadsheet, depending on the uptake… but independently managed – by the MRS, Esomar, ASC or SampleCon.

So please come to ORT House in London, on Thursday the 9th of November. Tell me who from your company is ready to speak and take part in the panel’s panel discussion, and in a few lines, give me an outline of how you’d respond to our challenge on harmonising panel data and panel interfaces by Monday 2 October. We’re looking for original thinking, fresh ideas and practical answers.

Panel Providers of the World, Unite!

Askia party at Esomar 2017!

A quite substantial contingent from Askia will be attending the ESOMAR Congress on September 10-13 in Amsterdam. For those of you who plan to attend, we’ll be at booth #19 in the exhibition space, where we hope you’ll swing by to see us.

The other bit of significant news is that, true to tradition, we’ll be hosting a party immediately following ESOMAR’s Welcome Reception. Details are as follows:

When?

Sunday, September 10th from 9:00 PM until late.

Where?

In De Waag, an intimate bar/restaurant about a 7-minute walk from the ESOMAR Congress.

In De Waag

Nieuwmarkt, 4

1012 CR Amsterdam

Dress code?

Something, anything…

Welcome to the machine

The following is a transcript of a talk given by yours truly and Chris Davison from KPMG Nunwood at ASC’s One Day Conference on the Challenges of Automation in Survey Research on May, 11th 2017.

Introduction

We have entered the golden era of automation –in other word: make machines do things. At first it was repetitive and simple things – find duplicates in a sample list, copy that survey and substitute the word Coca-Cola by Pepsi and send the results to all the executives of the relevant company – not mixing up companies is the perilous thing that you do not want to get wrong.

Automation is for lazy people – and I have always considered laziness to be a quality! Lazy people [programmers] look to avoid doing things they don’t really have to, and when they do finally have to, they look to get it done with the least amount of effort.

Now if we are to believe claims automation can post-code open-ended responses (yeah right Tim!), write reports, win at Go (mark my words – that one is never going to happen) and soon Skynet is going to enslave us (it will be ok as long as we don’t choose the red pill… or the blue pill.. damn I have to remember which one).
Automation is not new. Software is automation. But not everybody is a programmer – well this is less true at the ASC.

For every-one to benefit from the work done by the best programmers in their sector, Application Programming Interfaces (APIs) were invented. An API is a pile of code (usually documented) inside a box accessible by another software without having to understand its inner working (or have access to the source code). Here’s a brief timeline:

Timeline

By 1999 it was cool to be a programmer (and not just at the ASC). Every web designer was now a programmer – writing awful code in JavaScript so they could animate their poorly designed website while growing and grooming their facial hair and sipping their Frappuccino soy latte.

XML was no longer the cool kid on the block JSON (JavaScript Object Notation) was – more compact, more elegant and directly usable in JavaScript. jQuery – a JS library, quickly replaced by Angular and then React – made it super easy to query any website and people started calling me a dinosaur because I am a C++ developer.

Automation was always possible before: to make software interact you needed a database accessible by both parts. Or start an executable from the command line… Or a file drop on an FTP server. All these are huge security risks. I am not saying that web services are not security risks – the risks are just less understood so easier to sell to your CEO.

So what was new in the survey world? Well everything.

Panel providers created APIs. First Cint then Lucid which lead to an explosion of DIY research. Software providers opened-up their APIs and some even documented it. I will not give you the list but these days, even SurveyMonkey has an API.
And for me the revolution came with CRM system – like SalesForce, Microsoft Dynamics or ZenDesk – opening up the Enterprise world. You could interview any customer after any touch point… understand what’s happening and adapt quickly, it’s called by one of our competitors the experience gap.

Behaviour is now captured outside of surveys. The “What and When” is known. Surveys can concentrate on the “Why?” and the “What if?”. Verve is managing insights for Walgreens, the owner of Boots. Thanks to the loyalty cards, they know you have bought paracetamol on the Monday, Ibuprofen on the Tuesday and nothing on the Wednesday (with an app and iBeacons they also know where you have walked) – and they can sample you accordingly and interview you to understand your buying pattern.
So the game is about asking the relevant question at the right time. In my opinion nobody does it better than KPMG Nunwood and that’s because – automation looks like magic when it’s well done – there is a wizard in the backroom somewhere in Leeds…

Case study: Nunwood (Chris Davison)

So I’d like to tell Nunwood’s automation story, how it started, the obstacles we faced and where we are now before later coming onto some of the ideas that we implement to take us further…
How did we start? Well I wish I could say we had some far reaching vision but in all honesty this is what kicked things off for us…

In case of emergency, panic.

Sadly there wasn’t much long term vision associated with our first foray into automation, that came later. What got us moving was necessity. We had several UK projects using bits and pieces of automation techniques but they still required manual intervention at key points to move the process along.

As we started to expand globally, we were faced with the challenge of processes needing to run at any point in the day – including when our UK team were tucked up in bed. Our first attempt at a solution was a successful failure – we met the needs of the project, but the mish-mash of command line, SQL scripts and Askia was very complicated and wasn’t very accessible to the entire team.

If we were going to extend the automated approach to other projects, it was clear we would need different tools and this is what brought us to LoadIt. While a simple tool to use, it allowed for a great deal of complexity meaning new starters could get to grips with it quickly but the more seasoned DPers could still deal with our most demanding projects. Later, its extensibility would allow us to integrate with our in-house developed systems such as our Fizz online reporting platform.

robot

Given LoadIt’s existing integration with Askia and the automation capabilities within Askia we soon developed the long term vision that we were missing – the Zero Hours Project.
Given certain conditions – mainly stability in the project’s design and outputs – could we automate all elements of data collection and delivery?

It was an ambitious goal and one that had to have some compromises – there would always be some elements that would need manual intervention, so “zero hours” really meant “very few hours” but that didn’t quite have the same ring to it.

Discussing these kinds of developments with the whole team raised justifiable concerns that this would mean people’s jobs, but automation does not necessarily mean reducing head count and it certainly wasn’t our goal.
The tasks that lend themselves well to automation are the ones that don’t change – removing these from the team’s workload would free up time for things that required the skills for which they were hired (primarily their problem solving abilities), the skills required for automation were also different from the typical work, meaning it provided an opportunity for people to learn more broaden their skillset. We could also expand the remit of the team – in particular we took on more responsibility for the configuration of our reporting sites. From an operational perspective it would mean we could go some way towards flattening out the peaks and troughs that had developed in our working patterns – driven by the fact that most of our work was tracking studies that came out of and went into field at very similar times.

Framed like this, it was a very positive message for the team and I’m pleased to say everybody got behind the idea.
The main challenge from the wider business was around quality control. If machines were doing all the work, who was checking it?
It was a valid concern but one that could still be approached with automation at the forefront. All surveys are based on a set of rules, however complex – each question has criteria that must be met before it is asked, so it’s reasonably simple to check that the rules have been met.
We created VB scripts that could test the rules and give a pass or fail to a set of Excel tables, this meant that the same files we were using for automated checks could also be verified manually passed to our Insight team should they want to double check things.

Quality stamp

So back to the original question – could we run a zero hours project?
The answer wasn’t simple – Yes, if you considered the caveats – when things didn’t change and we considered the standard elements of the project we could produce the outputs through automated scripts. No, because an unexpected consequence of the changes was a change to the way our Data Team worked with our Insight Team. With many of the repetitive, standardized tasks removed, we found we had more time to work on ad hoc requests and deeper analysis of the data – meaning we could add more value to projects.

We had seen many of the improvements we had hoped as well as some unexpected ones: operational improvements, better working practices and we’d started to extend our capabilities.

Paradata (Jérôme Sopoçko)

One of the most exciting areas of using automation and APIs is during (or just after) the collection of the survey. Paradata is often just the date and time of the start of the interview. But more generally it’s about storing any information about the way the interview was conducted. You can find out the name of the browser, the operating system and the language used in the HTTP request header.

If the interviewee is using Internet Explorer 5 (or more generally any version of Internet Explorer), do not bother asking technical questions. Similarly, if the operating system is Linux, forget these ask technical questions because you won’t understand the answer.

If IE is brave enough to ask to be your default browser...

Beyond the interview, you can find information about the world. If you interview someone who is boarding a Eurostar train, it’s interesting to check the volume of #Eurostar hashtags in the Twitter API: it’s a strong indication of problems on the line.

Now let’s talk about the IP address – this identifier assigned to you by your Internet Service Provider. Of course your ISP knows who you are and has been allowed to sell (in the US) your browsing history.
There are a number of companies out there who specialise in transforming an IP address in a geographical position: www.freegeoip.net, www.maxmind.com, www.digitalelement.com, …

But you can have a much better definition of the geo-location by authorising the browser. Google thanks to their StreetView vans was fishing for the Wifi network information so they can improve the location which is now at scary levels of accuracy.

As mentioned to have accurate geo-localisation, you need the permission of the user. The idea is not to gather information in a sneaky way. Tell the user what you are doing with this information, explain that you are reducing the number of questions you ask… Because there are so many resources available: openweathermap.org will give you the current weather in any location, developer.zoopla.com will find right away the average price of the house in the vicinity. And then you have the open data government sites. Data.gov.uk have put on 185,000 datasets. Call me old fashioned but I still think we are in Europe, the EU Open Data portal has 10,700 dataset with a full API to access them. For free.

So what can we do with this data? Does that help to know that my interviewee is in Cardiff where 60% of the people voted remain? Linking big data and survey data is one of the greatest challenge of #MRX – and if you are not one of GAFA (Google Amazon Facebook Apple) you are at a real disadvantage.

Let’s use a concept developed for advertising planning: the Average Issue Readership (AIR)
We ask a significant amount of people (very significant in the case of the TGI survey) these questions: “How often do you read this newspaper?” or sometimes rephrased in “When did you last read this newspaper?”. There is still a lot of discussion to find out which is the best way to ask these questions – they are usually called probability questions.

Grid question

So you get a very classic grid question like the above. Thank to these lovely people who do research on research, we get the following probabilities of reading, based on the “recent reading” question.

Recent reading - NRS survey

From there we can infer, for each interview, the probability that he has read or not any given issue of a newspaper… and the great thing is that we can use that information in crosstabs and crossing that by their likelihood of buying Corn Flakes and plan our advertising campaign according to this.
In a normal survey, if we simply ask people what newspaper they read and we cross that by their gender. When we do a cross-tab, for each interview of gender who has given brand Y, we add 1 in the corresponding cross-tab. With probability question, we add a value between 0 and 1 indicating the probability of the person having seen the brand.

AskiaAnalyse table with randomised data 01

AskiaAnalyse table with randomised data 02

OK, the data looks a bit weird because the counts have decimals but once you move them into percentages, nobody cares anymore. And of course you can still use weighting if your panel data was not balanced.
So what does that mean for you: although you never asked for the vote that fateful referendum, you can cross the NPS of your brand with the vote for leaving Europe who (from what I have seen at the MRS) increasingly used as a segmentation tool.
Of course it’s not just the election you can cross by: the level of crime, the amount of subsidy, the likelihood of rain…

Beyond paradata, you can also create additional information with the questions you actually ask. We have worked on a project where the conjoint analysis utilities were computed in real time – that meant automating R (like Ian showed earlier) to get the results a few screens further, show the best concept for a given user and validate it.
Beyond that – the revolution is also around open-ended question analysis: you do not write open-ends anymore.

You will speak to your device, your computer, but also your phone, tablet, Alexa, your fridge…any IoT device – which have ways of recognising you. MyForce, our sister company, works on Bison – a revolutionary platform of not just speech to text – but identifying people by their voice (who’s talking), classifying the tone and talking speed (how are we talking) and the content (what are we talking about).

It’s not just Bison – look at what APIs Microsoft offers (Microsoft cognitive services) and R is integrated in SQL Server…

Microsoft APIs

Google and Facebook are also on the bandwagon (the gravy train).

One of our clients, through our other sister company Platform One – Nuaxia – has a panel of 1,000,000 doctors (not all in the NHS). These guys are in a hurry but they have interesting things to say. So, Nuaxia lets pharmaceutical labs survey these guys but only ask them 10 questions. And instead of asking them to type, they film them.

Platform One interface

This is the interface to create the survey – this is kept simple – it’s for pharmaceutical people. From there a survey file is created through the API of a well-known software vendor, they debit the PayPal account of the white blouse guys, invite the doctors and the data pours in.
After that, the video data is nicely sent to a speech to text algorithm, the text data is classified with Artificial Intelligence (a la CodeIt but not CodeIt yet) and all of it sent to a dashboard.

Text-driven surveys (Chris Davison)

So we know what the typical survey is structured like and most have not moved on that much from the sort that would be posted to someone in the distant past. Linear structures and the logic dictated by closed questions. Technology gives us an opportunity to flip this paradigm on its head.
Imagine a survey more like this…

What we’re looking do is use open end questions to determine the route the customer takes through the survey, asking things that are relevant to them and providing a much more tailored survey experience.

Removing the structure from our surveys is, for me, an exciting proposition and live text analysis can be used to do just that.

Create a pool of open-ended questions and as one is asked, apply live text analysis to determine which would be the most appropriate follow up and continue until either there are no more relevant questions or some constraint, such as time limit or number of questions has been reached.
From a respondent’s perspective, they should have greatly improved experience – far less asking them questions that do not seem relevant, the questionnaire is steered by the issues they want to talk about.
From the analysis side, the data quality should be much greater – in theory you’re asking questions of the respondent that are relevant to them and their experience. Consequently the ability to understand the story behind the data should also improve.

We can also start to tackle some of the issues facing us such as falling response rates – when an invite says the survey will last 10 minutes we can guarantee that – once the time limit is reached stop picking new questions. Or take a different approach, state the number of questions you’re going to ask and don’t ask anymore.
You can always ask the participant’s permission to ask more when you get to the limit, but because you’re asking the most relevant questions to them you hopefully have got the most interesting feedback up front.

There are clearly some analysis considerations – by only asking people about topics they’ve expressed an opinion about could introduce some bias, but nothing about this approach precludes randomly selecting questions or sections to provide balance. But you know when you’re doing that – you know the context in which the question was asked when it comes to analysis – you can even tailor the way it’s worded…
“We know you didn’t mention anything about your experience at the checkouts, but we’d like to ask you about it…”

To take this a step further you can then allow participants to upload photos / videos and do the same real-time analysis and base the survey route from that.

So while this is a specific example, the key principle for me is that we start to utilise the technological landscape we have available to us to start to challenge some of the fundamentals of project design. Connecting through the myriad of APIs helps us to create a combination of services that moves our industry forward and opens up new horizons.

Fresh articles from our knowledge base

Our tech support team has been hard at work lately on providing you with many new articles designed to help you better benefit from our software suite: could it be survey design, fieldwork management or data analysis.

Here are some of the most noteworthy articles from our help centre:

Randomise or rotate a code list with groups & headers

It’s often required to show a code list with several group headers and responses within these groups. In most cases it’s necessary to randomise the responses within the groups and / or the groups themselves. This article shows how this can be done using the Change response order routing action providing and example .qex and syntax. This functionality is often referred to as Block randomisation.

List of response items with response headers

By combining the use of a special character and of the Change response order routing to identify the response headers, you can easily setup your block randomisation as seen below:

Block randomisation

Take a look at the article for more information.

HTML table around screen elements

It can sometimes be hard to layout elements exactly as you want on a web screen if there are many elements or loops in play. This article will step through neatly storing several elements of a loop on a merged screen in a HTML table.

HTML tables may seem outdated but they are still the best way to semantically layout tabular data on web pages. Such layouts as the one below are nearly always best designed using the HTML table element:

4 question loop with a table layout

Check out the help centre article for more details and access to an example survey.

Using Google’s reCAPTCHA in AskiaDesign

The main purpose of integrating a reCAPTCHA in an askiadesign survey is to protect your survey from spam, abuse or bots while letting your respondents pass through with ease.

The only requirement is that you register your domain with Google’s reCAPTCHA in order to obtain a private key and add the appropriate JavaScript snippet to your survey.

You can play with this on our demo survey now or head straight to the KB article!

Statistics on mailing

This short article provides a description of each category located in the Statistics on mails tab in askiafield’s Supervisor.

Statistics on emails

Read the full article here for more information.

AskiaField 5.4 is ready for you

We are thrilled to announce that Askia’s entire software suite is ready for deployment in version 5.4.

Among the list of features of this major version, we want to highlight the very reason that made us switch to a new version track.

Quota revamp

quota54

5.4 introduces a total revamp of the quota system. Among the main features are

  • Quotas can now be set up on multi-coded questions and numeric questions.
  • Minimum and maximum quota targets.
  • Easy crossed-quotas.
  • Quota definition, monitoring and quota breakdown by interviewing mode are now on one single view.
  • Copy-paste large quota structures from Excel.
  • New keywords to master least-filled setups with a single routing.

This means a lot more control during the crucial moments of the fieldwork : balancing the last interviews so that all quota targets are filled up together, neatly, and fast.

You can check all new quota specs in the Knowledge Base.

And for the curious crowd, here is a good read if you’re interested in discovering what’s behind the scene, from a previous blog post.

CAWI to CATI, back and forth

5.4 is our first fully bi-directional multimode version. You can now natively switch interviews from CAWI to CATI and vice versa.

multimode

Field API

5.4 lays the ground stones for extensibility with Field API, allowing you to build ambitious end-to-end automation systems. Upon checking an extra checkbox during the setup, AskiaField can now receive orders from the outside. This also leaves an open-end for various integrations (with CRMs, CMSs, productivity systems….you decide).

Here is our developer reference if you’re interested.

Getting started

Like what you see ? Check out our full KB section to start mastering your new toolbox, and contact our support team to schedule the installation