Q Blog

Splitting out variables can get your game rolling

Matt — Fri, 21 Oct 2016 19:03:00 +1000

A bit like pool balls in a triangle, you may be aiming to split them apart.

When Q sets up your project, it will put alike variables together into a Set Question. This is useful of course, but sometimes you do in fact want to use individual variables within a set question.

There's a few ways to do this - one way is by selecting all the variables in the variables and Tab, and then copying/pasting the variables (exact copy). Another way is via "reverting to source".

But a convenient way, is the handy feature when you right-click in the Variables and Questions tab - it's called "Split variables from question". As the name suggests, it splits out all the variables from the question for you, so they are each ready to analyse as their own Pick-One question.

It's not just a useful trick for Pick-One Multi's, it can be used as well with Number-Multi's (to create a series of Number questions).

Give it a try today with one of your "Multi" questions.

For more information, see our Q-Wiki.

Save the date!... with date questions

Matt — Fri, 14 Oct 2016 18:42:00 +1000

Date questions give power to your tracking studies

When Q sets up your data, it sets up questions based on the structure of your variables. If Q can recognise a date(or date-time) variable, it will set its question type to Date - which has this little question icon:

Date questions can be very useful, particularly for tracking studies.

In tracking studies, many researchers use a single categorical (Pick-One) question to denote their time period (eg: wave 1, wave 2, quarter 1, quarter 2, etc).

But using a date question gives you functionality and power beyond what a single categorical question can give you. For instance, you can instantly change the time-banding of a date variable (eg: change it from monthly to quarterly, or vice-versa).

In fact, date questions offer many benefits - too many to cover fairly in this post in detail! But as an overview, these benefits include:

Changing between time periods with a single click (just going to the question's Values)

Unlocks a new statistical feature - "compare to last period" (very useful for trackers!)

Creating moving averages and other "smoothers"

Dynamic filters (eg: you can create a filter of "the most recent quarter" - which will update when you update the data)

Used in the periodic recomputation of weight variables

For more information on date variables, see the page on our Q-Wiki.

Filter one question by another... creates a new question!

Matt — Fri, 07 Oct 2016 20:02:00 +1000

A very useful QScript applicable in a variety of filtering scenarios

Sometimes we want to criss-cross two questions. There are lots of examples where this might be the case. For example, rather than applying a filter in a table we may want the filter to be "built-in" to a question (sometimes this can be used to rectify a situation where missing values haven't been encoded properly in the survey).

A very useful QScript is Filter One Question By Another Question. This can save oodles of time by avoiding crafting JavaScript or Logic variables.

It can be found in the Automate Library under "Filter".

Running this QScript will get Q to ask you two questions:

A) What question would you like to filter?

B) What question would you like to filter by?

And the result is a new question. Q even labels it as "Question A - filtered by Question B", and this new question appears in your drop-down menu(s). Q prompts you as to whether you want to hide the original question A and question B (as you may not need to see them anymore).

It's a very flexible QScript, in the sense that it can handle a variety of question types (ie: Question B doesn't need to be an official "filter" (binary) variable - it can be any question).

Full details on the QScript and how to run it, can be found on the Q-Wiki.

Automatic hiding and unhiding

Matt — Fri, 30 Sep 2016 06:49:00 +1000

Tuck away empty rows and columns with Q's automatic hiding feature

Sometimes rows and columns have no data. Perhaps this is because no counts are recorded against a particular option (maybe it wasn't a popular choice, or perhaps because of skips in the questionnaire).

Hiding can be done manually on the labels of a question with a right-click (there's a "Hide" and "Unhide" option) - you can do this on both the rows or columns of a table.

But you may want to hide ALL the empty rows/columns in one go. And furthermore, sometimes you may want this to happen automatically (so that if the content of your table changes, then the hiding and unhiding happens without you having to do anything else).

So Q built a convenient short-cut feature for this very reason - you can hide and unhide empty rows/columns very easily, with a single click, and that updates dynamically if your table changes.

The shortcut is right under your nose! (many don't realise it's there).

Once this button is toggled for a particular table (or group of tables) any empty rows or columns will be hidden, all at once.

And if you want to turn it off - simple! Just push it again and it will stop the hiding/unhiding.

Give it a try on a table with empty rows or columns!

Two further points:

This ability also lives as a Rule in the Automate library
The hiding ability can conflict with some statistical tests - if you encounter this problem, you may need to remove the categories manually to alleviate the conflict

The simple way to change how a question looks

Matt — Fri, 23 Sep 2016 06:24:00 +1000

The right-click is your friend when manipulating outputs

Inevitably as we analyse data we often need to change the way a question looks. This might be a recode of some values so we can add and average to a categorical question, merging categories to summarise a question, toggle missing values to change the sample base or change the question type to do different analyses or tidier crosstabs.

Despite the name, the Variables and Questions tab is not the only place in Q that you can manipulate your variables and questions. You can in fact directly manipulate questions from the tables tab.

Your friend here is the right-click.

Right-clicking on a table gives you lots of options to change and transform the questions. This includes:

Duplicating the question
Changing the values
Removing categories (which is the same as setting a category as a missing values)
Renaming a question
Transforming the question type
And much more!

And you haven't had to go to the Variables and Questions tab at all!

For example, to change the question type (as show in the picture above):

1. Right-click anywhere on your table

2. Towards the bottom, you will see your Question(s) (by their label)

3. Follow the selection path under your question to "Restructure Data"

4. Choose the new question structure

Your table will then be instantly updated to reflect the new question type.

(Further tip: Use right-click to make a duplicate of your question - then when you modify your question, you can keep a copy of the original. Don't worry, you can also any question to its original with the revert option too!)

Sneak peak at Q5 with full R integration, new regression, better PCA, web scraping and more

Susan — Mon, 19 Sep 2016 16:33:00 +1000

Last week we showed off some of the new features in the soon to be released Q5. Featuring a huge overhaul of advanced methods, including a host of new regression features, better PCA, and thousands of additional functions now available via code. Q5 introduces the ability to run code in the R Language from within Q. Major changes are detailed here: http://wiki.q-researchsoftware.com/wiki/Q5_Key_Features

How to find the most interesting crosstabs - instantly

Matt — Fri, 09 Sep 2016 21:09:00 +1000

Smart Tables will find the story in your crosstabs

You have a mountain of data and you need to find a story. What is interesting, what is not. Manually sifting through hundreds of crosstabs is too time consuming and error prone.

Try Smart Tables. It is a tool that automatically generates interesting crosstabs and ranks them in order of interest. Like surveying a landscape – Smart Tables help you see the peaks and troughs.

Here's what you do:

1. Decide on a key question (e.g., brand preference, overall satisfaction)

2. Select Create | Smart Tables....

3. Select your key question in the blue drop-down.

4. Click on any or all questions in the list and then hold down Ctrl and type A to select all the questions.

5. Click on the arrow to use the selected questions as Independent Questions.

6. Click on OK.

A series of tables will be generated in your report tree, ranked by their overall significance (p-value – the lower the p, the more “significant” the effect).

Give it a try. You will save oodles of time!

More details about Smart Tables are on our Q-wiki: http://wiki.q-researchsoftware.com/wiki/Smart_Tables

Q Research Software appoints 8 new staff

Susan — Wed, 24 Aug 2016 10:50:00 +1000

Numbers International, developers of Q Research Software, DataCracker and a soon to be released data science platform has made 8 new hires in the USA, UK and Australian offices.

Founder, Tim Bock (pictured), says “We are constantly innovating in a market that is hungry for modern analytical tools. The next month will see R fully integrated into Q and later this year we will launch our data science platform. So the expansion will continue”.

US based Tim Ali joins the support and development team. Tim has a strong technical background, both in terms of survey research and business intelligence.

Experienced quantitative researcher and former Kubi Kalloo consultant, Matthew Steele, joins the UK team to support brand and business development.

Mattias Engdahl, from DataSquirrels, brings substantial data processing and market research experience to the Customer Engineering team. Carmen Chan having just completed her PhD in Bioinformatics also joins the Customer Engineering team.

The development team welcomes three new members including Po Liu and Michael Wang who will use their solid software engineering experience to focus on data visualisations. William Guo recently joined the Sydney team and is responsible for all things finance related.

Q Research Software Opens Offices in London

Tracy N — Tue, 11 Nov 2014 15:24:00 +1000

Q Research Software has expanded operations to London. The London office will be headed by Nigel Cartman, a former Director of Information Tools (Europe).

Cartman says: “I expect that Q will be the next industry standard in the UK. It is without question the strongest and most flexible software I have used for market research data analysis and modelling”.

The new offices will provide sales and technical support to Europe. Q usage has grown rapidly in the UK, and it is the leading market research analysis software in Australasia.

Q research software is developed by Numbers International, based in Sydney Australia. Launched in 2007, Q enables market researchers to rapidly analyze and report data using the smartest statistical techniques.

Q is fundamentally different from other statistical programs because it uses the data file to automatically choose the appropriate analyses and statistical tests. This makes Q easier to use and produces higher quality (and often more interesting) results much faster.

To book a demonstration and free trial: email nigel.cartman@q-researchsoftware.com or visit www.q-researchsoftware.com

UK address: Hub Westminster 1st Floor, 80 Haymarket, New Zealand House, London SW1Y 4TE PH: +44 77 3993 3153

Q Research Software opens in New York

Tracy — Tue, 09 Apr 2013 22:16:00 +1000

Sydney based Q Research Software has expanded operations to New York. Opening later this month, the New York office will be led by former Illuminas Research Executive, Steven Kawalek.

New York based Kawalek says: “I am excited to see how researchers here respond to Q – I feel like I am about to expose one of the best kept secrets in the US market research industry”

The new offices will serve as a sales and technical support hub for the region. Q usage has recently grown exponentially in the UK with steady sales in the US. It is the leading market research analysis software in Australia and New Zealand.

Marketing Director Matilda Sarah says: “We were very lucky to find Steve. He is a skilled researcher, a seasoned Q user and he has a great understanding of the US market.”

Q research software is developed by Numbers International, based in Sydney Australia. Q was launched in 2007.

Matilda Sarah will be in New York from 29th April to meet with current and prospective Q users. For more information visit our website www.q-researchsoftware.com or contact us at matilda.sarah@q-researchsoftware.com or steven.kawalek@q-researchsoftware.com

The New York office address is: 116 West 23rd St, 5th Floor, New York, NY 10011.