# Protobi blog

We just got back from CRC2018, the Corporate Researchers Conference in Orlando Florida.

As much fun as it was speaking with research firms about our visualization platform, it was exciting to meet other innovative firms and see possibilities created by their products.

Speaking with a few of our colleagues about our favorites, it turns out that not every researcher knows about them. Here are a few we think are particularly interesting:

Flow diagrams can be a good way to visualize relationships between variables, like progression of treatment regimens by line of therapy.

One type of flow diagram is the Sankey diagram where the width of the arrows is proportional to quantity. Here’s how to create one in Protobi…

Ever review your data and wonder “What?! How did I get a mean of 2.13 on a 2-point scale?”

Surveys sometimes code special values like “Not asked” or “Don’t know” as integers like 9, -9 or 99. These can definitely throw off your analysis.

Here’s how to fix them in Protobi…

Sometimes your data has outliers. Trimming and Winsorizing are two ways to mitigate the effect of extreme values on your analysis. Two more alternatives are to recode or simply retain them.

Coding verbatims into concepts is a common task in text analytics. But how many concepts should you expect to find given your sample size? How big should your sample be to identify 20 concepts?

That may sound abstract, but when budgeting research that’s the bet we make with actual dollars. It’d be good to know the odds.

This article suggests a new way to predict how many distinct codes you may expect to see in N survey responses. Such a curve might be used to inform sample size selection before fielding research, or during analysis to benchmark the results.

The Van Westendorp Price Sensitivity Meter (PSM) is a non-parametric chart used to summarize stated consumer price preferences. It allows product managers to see the intersection between prices customers perceive as good value versus prices customers perceive as expensive.

Here's how to create it in Protobi using cumulative line charts...

Your survey data might have one or more columns with date values. There are lots of ways you can parse and analyze dates in Protobi.

How do we describe the distribution of time intervals when some aren’t yet complete?

The Kaplan–Meier Survival Estimator is a non-parametric curve that describes the empirical survival function given observed interval to-date.

Importantly it is designed to handle “censored” data where the intervals are observed before they are known to be complete.

Surveys often ask for time intervals, with start and end dates:

• When did you buy the product? When did you finish it?
• When did the patient start and end each line of therapy?
• When did respondents start and end different programs?

One thing we can do is to look at the data. Another is to look at how survival data is summarized in clinical research…

Surveys can contain “loops” where a subset of the survey is repeated several times per respondent. This is typical in new product assessments, employee satisfaction surveys, patient case research, and observational trials.

You can choose whether to see survey loops “flattened” or “stacked”. Which is best depends on your analysis goals.

We're excited to support SERMO Dashboard Analytics! Protobi Viewer is now available with every SERMO RealTime and full length survey globally.

See the intro video:

Your survey asked quantities as absolute counts. But now you need to report them as percentages. Here’s how to calculate ratios and correctly preserve percentages, frequencies and means:

Yay! You’ve fielded a global survey in multiple local languages.
Yikes! Now you need to analyze all those local-language verbatims…

Protobi works with Google Translate so you can start reading and even recoding those text verbatims in multiple languages to analyze right away.

Straightliners. You know they must be somewhere in your sample … respondents who give the same answer to every question in a section.

If you could see the answers for one respondent for one section, it’d be easy to spot. But how do you quickly identify all straightlines? It’s pretty easy to find them in Protobi using this one trick…

As you work, Protobi saves all your changes locally, and your latest version survives closing the browser. You can work on your own copy and push changes up to the server when you’re ready for colleagues to see. Work from an airplane or ferry, then sync your changes when back online.

Select “Local History” from the toolbar context menu (or press Shift+Z). You’ll see a timeline of your most recent changes. Select a timestamp to restore your project as it existed at that moment:

Interactive analysis is great for exploring the data, testing hypotheses. Collaborating online is great for finding the story with colleagues and clients. But in today’s business world, analysis still has to go into PowerPoint to tell that story to the broader organization.

Protobi lets you create visualizations that look more like your presentation than your survey. And export into your own PowerPoint template as editable chart objects.

Dynamically resize any chart in Protobi with the mouse. For any selected element, a resize handle appears when you hover.

You can show pretty much any distribution as a WordCloud. For instance, you can show the states where survey respondents are located:

Perceptual maps can be a useful way to concisely visualize associations among multiple variables. Protobi can create a perceptual map based on principal components analysis for many types of crosstabs.

Create Wordle-style word clouds in Protobi for text verbatims

You’ve asked each respondent to answer multiple questions. Now you want to know if respondents’ answers to this question are significantly different than their answers to other questions.

Protobi’s new PairedTable allows you to compare different questions across the same respondents (rather than compare the individual questions for different subsets of respondents). This uses pairwise comparisons for stronger statistical tests.

This uses pairwise comparisons for stronger statistical tests. It uses pairwise t-tests to compare means and McNemar’s test (with small sample corrections) to compare percentages.

A TopBoxTornado plot is a concise way to present top- and bottom-box scores for multiple ratings on Likert-type scales.

Protobi is not just a pretty face for the data, it also provides a full-featured language for data cleaning and reshaping prior to analysis.

No matter how carefully you design a survey there are almost always changes you need to make to the data once it comes back:

• combine multiple waves of an ATU
• merge in translations for text open-ends in other languages
• stack patient cases
• calculate time intervals
• define segmentations
• remove outliers
• zero-fill skipped values

You can now do all of the above (and more) within Protobi.

Previously you might have used SPSS, R, or external vendors to do this externally. You can still do that, and upload the results to Protobi as you wish.

But now you can also keep all your processing code in one place, integrated with your analysis, and documented.

Strapped for time? We’re happy to set up your data cleaning and reshaping for you, and show your analysts how to edit or author it.

Back-to-school season entails all the necessary checkups and health exams.
Seeing where the kids fell on the height weight standards chart, I noticed that the charts all seem to conveniently stop at age 20. What would they look like if extended for adults?

We love bar charts and their simple utility in the New York Times and Wall Street Journal. But other chart types also have their role in finding and telling the stories in survey data, and our client work often entails creative custom visualizations…

Does your survey include a collection of related questions on a common scale? E.g.

• Ratings: “How strongly do you agree with the following…?”
• Frequencies: “How often do you do the following activities…?”
• Rankings: “Please rank these items from most desirable to least…”

Protobi includes useful tools—top-box summaries, stacked bars, crosstabs and clustering—that make it easy to analyze ratings, rankings, and other questions on common scales. But the tips here you can do in Excel, R or even PowerPoint…

(hover to expand)

As powerful and ubiquitous as the mobile/web has become, PowerPoint is still the platform for business analysis today. Interactivity and rapid collaboration are awesome, but business findings are presented in slides, to be presented, distilled and synthesized, as insights crystallize into decisions.

So we’re pushing the boundaries of PowerPoint on the web, making it easy to export data as slides with native charts and tables. Using your company’s template. And even to instantly update the charts and tables (leaving your text untouched) as new respondents come in.

Which got us to wondering, if all other business results must be presented in PowerPoint why don’t executives ask to see the New York Times in a slide deck? Maybe just no one ever thought it was possible!

So for fun we combined the NYTimes Top News API to test our shiny new library to generate PowerPoint with native charts, images, tables, real time data and user-defined templates.

More to the point, Protobi can export your entire survey… be it in Survey Monkey, TypeForm, Qualtrics and Confirmit Surveys seamlessly to native PowerPoint charts and tables.

Wait! Before you send the survey for programming, here are a few quick ways to simplify the survey for the respondent, the client, and the analyst.

A few of these pick on an actual customer satisfaction survey Amtrak sent me. This is unfair. The trip was great, and it’s clear that Amtrak listens to its riders to keep the experience nice. But a lot of surveys are similar so it’s a good example.

A key task in any survey is identifying outliers that can mar an otherwise great analysis. Outliers can arise for many reasons – honest mistakes, careless entries, or outright bogus answers. Protobi makes outliers stand out so identifying them as easy as shooting fish in a barrel.

How many ideas might you expect to find in customers' responses to open-ended survey question? Here's an interesting empirical analysis of text verbatim coding from a recent survey, looking at actual data compared to expected values under Zipf's Law and Heap's Law.

The survey question was "Why did you choose the product you selected?". Respondents provided free-text responses. 200 responses were coded in Protobi using the new verbatim coding widget by a professional analyst.

Verbatims from open ended survey questions are a rich source of insight for market researchers, and a great way for your survey to tell you something you didn’t already know. But surveys often don’t include them, as analyzing text responses has historically been a hassle.

What if coding text verbatims were fun and easy? Would we ask them more often? Might we learn more of what the market is often very willing to tell us?

If you have a current survey with text verbatim responses, let us know. We’re running a study you might be interested in…

How do you find the optimal price for a good or service? Obviously, it depends what you mean by optimal. And pricing is a hugely complex problem. But if your goal is narrowly defined to maximize expected profit based on a discrete choice logit model, this page has an elegant new solution.

We present a simple analytic formula for the optimal price in a discrete choice pricing model. Here, the optimal price is the one that maximizes the expected revenue (or profit), balancing the revenue versus the likelihood of purchase. This formula allows the optimal price to be quickly calculated precisely for each individual customer, for further analysis and action.

Protobi now enables drag-and-drop recoding for text verbatims!

Thanks to our users for awesome suggestions in a recent series of user labs! Key themes:

1. using Protobi to create client deliverables on rapid timelines and
2. presenting Protobi as a client deliverable.

This release introduces several new capabilities:

• Copy elements rather than just move them
• Find-and-center an element by double-clicking on it in the tree.
• Save scenarios in a new toolbar button
• Evaluate scenarios as a segmentation for crosstabs
• Define new segmentations logically using Mongo-style constraints.
• Define new segmentations functionally using Javascript expressions.

A frequent question prospective clients ask is “How are you different from Tableau?”

On the surface, Protobi and leading BI tools are similar in that both create clickable graphs and tables from data. Beyond that they’re radically different tools for different purposes, and even coexist quite nicely.

Was encouraged to participate in the MIT Big Data Hackathon at Hack/Reduce in Cambridge, MA this weekend by a friend Ashwini Kumar, principal engineer at Senscio Systems. The very idea of signing up to work into the wee hours amongst the super talented people one would imagine would be there seemed both intense and pretty intimidating. But he’d been to these before and assured they are really positive sessions from which you learn a lot you’d never expect. Plus my kids thought the idea was cool. So I was in. And wow, they were right.

“Hey, is there any way we can see this data as a perceptual map? I’d love to show this to the marketing team tomorrow.” Thus was the question posed by a smart client who likes to present synthesized findings rather than just raw data.

The next question was “How can we make this easier to see?” Together we worked out a new take on Perceptual Maps to make more fun…

You can now export Protobi visualizations as Excel workbooks with statistical tests. Set a crosstab banner and generate an entire deck of crosstabs. Use it for your own advanced analyses or send it directly to a client.

“This is great! How can I export visualizations to PowerPoint for my clients?” We hear you!

You can now export Protobi visualizations as PowerPoint presentations:

• Charts are native PowerPoint chart objects
• Data is embedded as Excel worksheets

We’ve updated the look of Protobi!

• Current distribution as a single solid color
• Baseline distributions as a thin black outline

This is based on end user feedback from user labs with our clients and theirs. As before:

• Blue triangles show statistically significant differences
• Gold bars show active filters

A key strength of Protobi is its ability to show statistical contrasts, comparing one subset to another. The design challenge is how to make this is clear, intuitive, and aesthetic. We think this strikes the balance.