The Effective Statistician - in association with PSI

Interview with Eva Murray and Andy Kriebel Click this link to go to the homepage of this episode! Today, I talk with these two awesome visualization experts - Eva and Andy. They have written great content about data analysis and visualization. Learn a lot about great visualizations from them. We discuss the following important points: How to create effective visualizationsHow to communicate this dataHow do people benefit from it References:MakeoverMondayWhat's Going on in this GraphChart Guide Listen to these awesome data visualization experts and learn from them!

An interview with Paula Willamson and Liz Gargon Click this link to go to the homepage of this episode! This clarifies directly what the COMET initiative is about. But it also leaves open a lot of questions such as: Why are COS important?What are the goals of COMET?How narrow or how wide should COS be defined?What does the process look like to get to COS? References: Homepage of COMETCOMET VIII ConferenceReferences from COMETCOMET DatabaseTwitterCore outcomes set in RA Listen to this episode of our podcast and learn from Paula Williamson and Liz Gargon about this important initiative.

Click this link to go to the homepage of this podcast! In this podcast, we talk about the following 9 important points: In this podcast, we talk about the following 9 important points: Overview of the ITIT courseWho is the course aimed atWhat are the reasons to attendHow the course keeps up with changes in the industryWhere does the course fit in with the organization's goalsWhy should a manager think about sending someone from their team on the courseWhich locations delegates can expect to visitWhat makes the course such a successFrom the delegates' perspective, what are the benefits for the companies hosting the sessionWhen does the course run and when should delegates apply References: Link to the ITIT coursehttps://www.psiweb.org/events/event-item/2020/10/01/default-calendar/introduction-to-industry-training-course-2016If you would like to host a session, please contact:Alex.Godwood@SoseiHeptares.com  or psi@mci-group.com Listen to this episode and learn from it!

Click this link to go to the homepage of this episode! And today, we are looking back and discuss the last 2 year's learnings. We discuss the 9 learnings we had and how they apply to you as a listener as well. Specifically, we talk about: Think bigger  Prepare for the unexpected  Delegate what you are not good at or what you don’t like  Have fun Relationships are the foundation for success  Be curious and dig deeper  Our community is big Unnecessary red tape differs a lot by company  Growth happens outside your comfort zone  We need to all step up to improve the impact of statisticians for the benefit of patients. Listen to this episode, share it with others, and learn more!

Episode 3 Click this link to go to the homepage of this episode! In episode 3 of this mini-series, I discuss the impact of simplified data visualization and how to achieve it. Specifically, I address these factors that affect your visualization: Clutter Mis-alignmentToo many visual linesExample misalinged header, legends, Long categories below vertical bar charts not fitting well under the bars (avoid diagonal)How many horizontal and vertical lines (start and stop of visual elements) do you have?Can you avoid a legend?Do you need these gridlines?Do you need tickmarks?Too many elements likeColourFontsBold, italicAnnotationsHeader, subheader, sub-subheader“Unnecessary” annotationsFile namesSource dataProgram nameValidation statusUnderstand gestalt principlesProximity - put the description for treatment next to the lines - not in a legendSimilarity - re-use color for the lines and the font color of the treatmentEnclosure- box around the most important area (e.g. randomized period vs open-label follow up)Continuity - bars being organized below each other have a same starting point - take out the line showing baselineConnection - connect what belongs together - line charts over timeClosure - take away the box around your overall vis Learn more on how to create simple but effective data visualization by listening to this podcast and share this with others who might learn from it!

Bonus Episode 2 - COVID Click this link to go to the homepage of this episode! Today, I'm speaking about Running on 2 legs - my call to action for statisticians to step-up in this COVID pandemic. How can statistician not just reflect on and analyze data, but to lead on letting the people understand the dataShort-term thingLong-term thingLet people understand statisticians' functions and role to societyStatisticians as strategic partners Enjoy listening to this episode and be more aware on what we can do to add value to this time of pandemic!

Interview with Alberto Cairo Click this link to go to the homepage of this episode! We also discuss the following points: How tables compare to charts in terms of “lying”How to strike the best balance between showing the details and overwhelming the audienceWhat behaviours or habits should we develop to improve our visualizations References: References: Michael Friendly's Golden Age of Visualization paper Eurostat's Power from Statistics about uncertainty,The story about the "New Golden Age" Tools: DataWrapper Flourish iNZight Books: Show Me the NumbersStorytelling Data Visualization Business ProfessionalsTruthful Art Data Charts CommunicationElements Graphing Data William Cleveland Visual Display Quantitative InformationExploratory Data Analysis John Tukey Listen to this episode and share this with your colleagues!

Episode 2 Click this link to go to the homepage of this episode! In this second episode. I talk about how you can get a good visualization. Here are the following important tips: Importance of detailsCo-creation with the audienceExample launch slide setKOL engagmentMultiple review roundsTesting and refinementExample trial level safety analysesIterative approach - validation level - documentation - fast turn aroundsStart with pen and paper and develop together fastFocus on the larger topics like which plot to use first and then fine-tune to a more detailed level (examples for each)Use of colourTitleLabels, legendsAlignmentFootnotesSupporting textDashboards (examples of each)Filter and sorting optionsArrangement of multiple displaysHighlighting optionsHover over optionsDigging deeper optionsEncouraging and giving feedback Listen to this episode and share this with your friends and colleagues who might benefit from it!

Click this link to go to the homepage of this episode! I discuss my learnings, my reflection about this, and what is the meaning of this situation for us, statisticians, and more significant factors like the following: Hardly any tables out there, but lots of visualizations both interactive and explanatory - still most of our day-to-day jobs are providing tablesTables are not the key deliverables episodeInterview with Zak Skrivanek - a figure says more than 1000 tablesWonderful Wednesday episodeVIS SIG Get the data from many different sources - not just the one source at hand: e.g. the study currently working on; RWE, literature data, and other studiesMany misunderstandings on the sources of the data and comparisons across different sources (e.g. different countries) and then infer on the differences between the policy making in the different countries - where are the statisticians in the news explaining the numbers?We need more leadershipOur associations need to step up and become more professional and impactfulWe have a responsibility hereResponsibility for providing the numbers but also gives the background and advise on how to use the numbers - for example John HopkinsFAQ not really helpfulWhat can be answered here and what notHow reliable is the data?What are the strength and limitations?Any guidance on the use of the data?Do we train people on how to read our data? Or do we just through the tables over the fence for someone else to deal with it?Power of scenario simulations - Washington Post article - great to show conditional probabilities e.g. if you have different prior information in Bayesian analyses Listen to this timely episode and let me know what you think!

Interview with Jürgen Hummel Click this link to go to the homepage of this episode! In this episode, Jürgen and I discuss about the following points: Why is working from home beneficial for the statistician and the employerWhat tasks are especially suitable for working from homeWhat tips do we have to make working from home successfulHow to balance office time and home office time Learn more by listening to this episode and share this link with others who might be interested in working at home!