Skip all navigation and go to page content

NEO Shop Talk

The blog of the National Network of Libraries of Medicine Evaluation Office

Archive for the ‘News’ Category

An OERC Resolution Realized

Friday, January 1st, 2016

Happy new year greeting card or poster design with colorful triangle 2015 shape and vintage label illustration. EPS10 vector.

Happy New Year, readers.  We decided to start 2016 on a high note with the OERC Blog’s Annual Report.

Wait, don’t leave! 

I promise, there’s something for everyone at the end of this post. There are links to the most popular evaluation tools and resources from our blog entries this year.  We judged popularity using our “clickthrough” statistics, showing the links that were most likely to be clicked by our readers to further investigate the resources featured in our blog posts.

(If you want, you can skip to the bottom now. We’ll never know.)

But first, we are going to take a moment to describe how far our blog has come since its inception in 2006.  That’s right, August 2016 is the 10-year anniversary of the OERC blog. I’ve been a contributing blogger for almost a decade. Karen, the newcomer, started contributing in February 2015, the month she joined the OERC. Her entries are quite popular: Most of the top 10 clickthroughs below were presented in Karen’s posts.

For most of the blog’s history, OERC staff posted 12-16 times per year (about 1 per month).  However, in January 2014, the OERC staff committed to increasing our blog activity. That year, we managed a little better than three posts per month.  This year, we finally met our goal of once-per-week posts.  We had 52 entries between January 1 and December 31, 2015.

Of course, writing blog posts is one thing. Writing blog posts that people read is another. Our first indication that our blog was gaining readership was through the OERC’s appreciative inquiry interviews conducted in late 2014. (See our blog post on October 31, 2014 for a description of this project)  Many of the interviewees mentioned the blog as one of their favorite OERC services.

Now, we have quantitative evidence of a growing readership: our end-of-year site statistics. Our stats only go back to June 2014. Even with that limited timeline, you can see substantial growth. Our peak month in 2014 had 241 total views.  In 2015, our peak month had more than double the traffic, with 508 views.  In 2014, we had an average of 7 views per day.  In 2015, our average was 12 views per day.

Two graphs showing increased readership for the OERC Blog. One shows increasing monthly views (June 2014=41; December 2015=452) Line chart 2 shows increasing average daily views (June 2014=7; December 2015=12)

So thank you, readers.  You are behind those numbers. In the coming year, we resolve to continue our weekly posts on a variety of evaluation topics.

And now, as promised, here are the top 10 “clickthrough” URLS from last year.  If you missed any of the trending evaluation resources from our blog, here’s your chance to catch up.

 

 

 

 

Fun for Data Lovers: Two Interactive Data Visualizations

Wednesday, December 23rd, 2015

‘Tis the season of gift-giving and who doesn’t love getting toys during the holidays? So we want to give our readers links to two fun data visualizations to play with over the holidays. Both were designed by David McCandless at Information is BeautifulSnake Oil Superfoods summarizes the evidence (or lack thereof) for health claims about foods popularly believed to have healing properties. Snake Oil Supplements gives the same scrutiny to dietary supplements.

You can readily check the science behind the infographics. Both link to abstracts of scientific studies indexed in reputable sources such as PubMed or Cochran Library. The pretty-colored bubbles and the filters give you an enjoyable way to check some of those food-related miracles being proclaimed in popular magazines and your Facebook feed.

Meanwhile,  Happy Holidays to you and yours from OERC bloggers Cindy Olney and Karen Vargas.

Christmas 2015
Cindy (left) and Karen at the American Evaluation Association’s 2015 Conference

 

 

 

Dashboards for Your Library? Here Are Some Examples

Friday, December 18th, 2015

illustration of different business graphs on white backgroundLast week’s blog post was about using Excel to make data dashboards. As Cindy pointed out a dashboard is “a reporting format that allows stakeholders to view and interact with program or organizational data, exploring their own questions and interests.”

What can that mean for your library? What does a library data dashboard look like?

In the OERC Tools and Resources for Evaluation, we have a Libguide for Reporting and Visualizing, which includes a section on data dashboards.  In it are some examples of libraries using data dashboards.  In their dashboards, libraries are sharing data on some of the following things:

  • How much time is spent helping students and faculty with research
  • What databases are used most often
  • How e-books are changing the library picture
  • What librarians have been learning at their professional development conferences
  • What is the use of study rooms over time
  • What month is the busiest for library instruction
  • What department does the most inter-library loan

Can you create a dashboard to tell a story? While libraries can keep (and post) statistics on all kinds of things, consider who the dashboard is for, and what story you want to tell them about your library.  Maybe it’s the story of how the library is using its resources wisely.  Or maybe it’s the story of why the library decided it needed more study rooms.  Or the story of whether or not the library should eliminate it’s book collection and increase e-books and databases.

Consider what data you want to share and what people are interesting in knowing.  Happy dashboarding!

Excel Dashboards at Chandoo.org

Friday, December 11th, 2015

Chandoo.org is a website that excels at Excel.  More specifically, it provides an extensive collection of resources to help the rest of us use Excel effectively.  There’s something for everyone at this website, whether you’re a basic or advanced user. Today, however. I want to specifically talk about Chandoo.org’s resources on building data dashboards with Excel.

Data dashboards are THE cool new data tools. A dashboard is a reporting format that allows stakeholders to view and interact with program or organizational data, exploring their own questions and interests. When the OERC offered a basic data dashboard webinar several years ago, we hit our class limit within hours of opening registration. If you are unfamiliar with data dashboards, here are slides from a presentation by Buhler, Lewellen, and Murphy that describe and provide samples of data dashboards. .

Tableau seems to have grabbed the limelight as the go-to software for data dashboard development. Yet it may not be accessible to many of our blog readers.  It’s expensive and, unless you are a data analyst savant, Tableau may require a fair amount of training.

The good news is that Excel software is a perfectly fine tool for creating data dashboards. Some of the best known data visualization folks in the American Evaluation Association (AEA) are primarily Excel users. Stephanie Evergreen of Evergreen Data  and Ann Emery write popular blogs about data visualizations built from Excel. At the AEA’s annual conference in November, I attended a presentation by Miranda Lee of EvaluATE on creating dashboards with Excel.  She has some how-to dashboarding videos in the works that will be available to the public in the near future. (We’ll let our blog readers know when they become available.)

Hand of a business man checking data on a handheld device

There are free resources all over the Internet if you are good at do-it-yourself training.  However, for a modest fee, Chandoo.org offers a more systematic class on how to design a data dashboard with Excel. Depending on how many resources you want to take away from the class, the cost is between $97 (online viewing only) and $247 (downloads and extra modules). I have not taken the class yet, but I have heard positive feedback about Chandoo.org’s other courses and have plans to take this class in the near future.

If you are an Excel user but don’t see dashboard-building in your future, you still may find a wealth of useful tips and resources about Excel at Chandoo.org. My favorite is this list of 100+ Excel tips. I attended several data dashboard sessions at the AEA conference last month. The word on the street is that Microsoft is rising to the challenge to develop its data visualization capabilities.  Apparently, each new release is better than the last.  It may be getting easier to work dashboard magic with Excel.

Measuring What Matters in Your Social Media Strategy

Thursday, December 3rd, 2015

Thumbs up symbols with text "get more likes"

We’re all trying to find ways to improve evaluation of our social media efforts. It’s fun to count the number of retweets, and the number of ‘likes’ warms our hearts.  But there’s a nagging concern to evaluators – are these numbers meaningful?

Your intrepid OERC Team, Cindy and Karen, attended a program at the American Evaluation Association conference in Chicago called “Do Likes Save Lives? Measuring What Really Matters in Social Media and Digital Advocacy Efforts,” presented by Lisa Hilt and Rebecca Perlmutter of Oxfam.  The purpose of their presentation was to build knowledge and skills in planning and measuring social media strategies, setting digital objectives, selecting meaningful indicators and choosing the right tools and approaches for analyzing social media data.

What was interesting about this presentation is that the presenters did not want to rely solely on what they called “vanity metrics,” for example the number of “impressions” or “likes.”  Alone these metrics show very little actual engagement with the information.  Instead they chose to focus on specific social media objectives based on their overall digital strategy.

Develop a digital strategy

  • Connect the overall digital strategy to campaign objectives: (for example: To influence a concrete change in policy, or to change the debate on a particular issue.)

Develop social media objectives

  • You want people to be exposed to your message
  • Then you want people to engage with it somehow (for example, sharing your message) or make them work with it somehow (for example: sign an online petition after reading it).

Collect specific information based on objectives

  • Collect data about social media engagement supporting your objectives that can be measured (for example “the Oxfam Twitter campaign drove 15% of the readers to signing its petition” vs. “we got 1500 likes”)

The presenters suggested some types of more meaningful metrics:

  • On Twitter you can look at the number of profiles who take the action you want them to take, and then the number of tweets or retweets about your topic.
  • For Facebook, the number of likes, shares and comments mean that your audience was definitely exposed to your message.
  • Changes in the rate of likes or follows (for example if you normally get 5 new followers to your fan page a week, but due to a particular campaign strategy, you suddenly started getting 50 new followers a week)
  • Number of “influential” supporters (for example, being retweeted by Karen Vargas is not the same as being retweeted by Wil Wheaton).
  • Qualitative analysis: Consider analyzing comments on Facebook posts, or conversation around a hashtag in Twitter.

Overall, your goal is to have a plan for how you would like to see people interact with your messages in relation to your overall organizational and digital strategies, and find metrics to see if your plan worked.

 

Take The Pie, Leave The Pie Chart

Wednesday, November 25th, 2015

Evaluation and data visualization folks may disagree on the best pie to serve at Thanksgiving dinner.  Pumpkin?  Pecan?  A nice silk pie made with chocolate liqueur and tofu? (Thank you, Alton Brown.)

You see, the whole point of charts is to give people an instantaneous understanding of your findings.  Your readers can easily discern differences in bars and lines.  In wedges of pie, not so much. Data visualization expert Stephen Few explained the problem during this interview with the New York Times: “When looking at parts of a whole, the primary task is to rank them to see the relative performance of the parts. That can’t be done easily when relying on angles formed by a slice.”

(Note:  This parts-to-whole angle problem may also explain why most of us can’t understand how our Whole Foods pumpkin pie could possibly have eight servings. Eight? Are you kidding me?)

So, for today’s pre-Thanksgiving holiday post, I thought I would point you to some online articles about the much used and much vilified pie chart.

First, here’s an article by American Evaluation Association’s president-elect John Gargani, arguing for retirement of the venerable pie chart.  He make points that are repeated in many anti-pie chart blog posts.  But in the interest of objectivity, you should know that agreement to send pie charts to the cosmic dust bin is not universal. Here’s a post by Bruce Gabrielle of Speaking PowerPoint that describes situations where pie charts can shine.

In general, most experts believe that the times and places to use pie charts are few and far between. If you have found one of those rare times, then here’s a post at Better Evaluation with the design tips to follow.

But for heaven sake, turn off that three-dimensional feature in your pie chart, or in any chart, for that matter. Nobody wants to see that!

And for humorous examples of what not to do, check out Michael Friendly’s Evil Pies blog,

Data Party Like it’s 2099! How to Throw a Data Party

Friday, November 20th, 2015

two funny birthday dogs celebrating close together as a coupleWhat’s a “data party?” We attended a program by evaluator Kylie Hutchinson entitled “It’s a Data Party!” at the AEA 2015 conference last week in Chicago.  A data party is another name for a kind of participatory data analysis, where you gather stakeholders together, show them some of the data that you have gathered and ask them to help analyze it.

Isn’t analyzing the data part of your job?  Here are some reasons you might want to include stakeholders in the data analysis stage:

  • It allows stakeholders to get to know and engage with the data
  • Stakeholders may bring context to the data that will help explain some of the results
  • When stakeholders participate in analyzing the data, they are more likely to understand it and use it
  • Watching their interactions, you can often find out who is the person with the power to act on your recommendations

So how do you throw a data party? First of all you need to know what you hope to get from the attendees, since you may only be able to hold an event like this one time. There are a number of different ways to organize the event.  You might want to consider using a World Cafe format, where everyone works together to explore a set of questions, or you could use an Open Space system in which attendees create their own agenda about what questions they want to discuss.  Recently the AEA held a very successful online unconference using MIT’s Unhangout that could be used for an online Data Party with people from multiple locations.

The kinds of questions Kylie Hutchinson suggested asking at a data party include:

  • What does this data tell you?
  • How does this align with your expectations?
  • What do you think is occurring here and why?
  • What other information do you need to make this actionable?

At the end of the party it might be time to present some findings and recommendations that you have.  Considering the work that they have done, they may be more willing to listen.  As Kylie Hutchinson said “People support what they helped create.”

 

The OERC on the Road at the American Evaluation Association Conference

Friday, November 13th, 2015

As you read this, the OERC’s Karen Vargas and Cindy Olney are attending the American Evaluation Association’s annual conference in Chicago, hearing about the latest and greatest trends in evaluation.  The AEA conference is always excellent, with evaluators from all disciplines sharing their skills, lessons learned, and new approaches to the art and science of evaluation. Look for our favorite topics from this year’s conference in our future blog posts.

In the meantime, you can get your Friday afternoon evaluation fix from the publicly available resources offered at the AEA web site. The AEA 365 blog has daily posts from AEA members that feature hot topics and rad resources on almost every evaluation topic imaginable.  Conference and other materials are archived in the AEA Public Library.  And don’t worry. The OERC will be back to regular weekly postings next Friday.

"Chicago sunrise 1" by Daniel Schwen - Own work. Licensed under CC BY-SA 4.0 via Commons - https://commons.wikimedia.org/wiki/File:Chicago_sunrise_1.jpg#/media/File:Chicago_sunrise_1.jpg
“Chicago sunrise 1” by Daniel Schwen – Own work. Licensed under CC BY-SA 4.0 via Commons – https://commons.wikimedia.org/wiki/File:Chicago_sunrise_1.jpg#/media/File:Chicago_sunrise_1.jpg

 

Infographics – Guide from NIH Library Informationists

Friday, November 6th, 2015

cool infographic elements for the web and print usageThe Medical Library Association’s October 28 webinar was on Data Visualization, presented by Lisa Federer, NIH Library’s Research Data Informationist.  The webinar was a tour of different aspects of data visualization, including information about elements of design, like color, line, contrast and proximity, as well as loads and loads of specific resources for more information.

For those of you who were not able to attend or would like to know more, Lisa Federer has a LibGuide called Creating Infographics with Inkscape, which contains the resources for a class she taught with NIH Informationist Chris Belter.  The LibGuide includes a Power Point from the lecture part of their class. The slides cover design principles and design elements.  Many of the slides have links to resources that you can use to learn more about the topic.  For example:

Vischeck – a cool tool for finding out what your colors in your chart look like to someone who is color blind

10 Commandments of Typography – suggestions for making font combinations that work

The second part of the class is a hands-on section on using Inkscape, a free, open-source graphics program, to make infographics.  Inkscape allows you to use “vector graphics” to design infographics.  What are vector graphics and why use them? You know images that work when they’re small but get all blurry when they get big? Those images are based on pixels. Vector graphics are based on pathways defined by mathematical expressions like lines, curves, and triangles, so they can get larger and smaller without losing any quality. Sounds hard to do, right? Luckily there are tutorials on Inkscape and it’s easier than you might think (you don’t need to know the math…): https://inkscape.org/en/doc/tutorials/basic/tutorial-basic.en.html

If you want to take a look at other vector graphics editors, there are other free ones, like Apache Open Office Draw, or ones you may already own, like Adobe Illustrator.  Comparisons with links to detailed information can be found in Wikipedia’s “Comparison of Vector Graphics Editors.”

Boosting Response Rates with Invitation Letters

Friday, October 30th, 2015

"You've got mail" graphicwith mail spelled m@il

Today’s topic: The humble survey invitation letter.

I used to think of the invitation letter (or email) as a “questionnaire delivery device.”  You needed some way to get the URL to your prospective respondents, and the letter (or, more specifically, the email) was how you distributed the link. The invitation email was always an afterthought, hastily composed after the arduous process of developing the questionnaire itself.

Then I was introduced to Donald Dillman’s “Tailored Design Method” and learned that I needed to take as much care with the letter as I did the questionnaire. A carefully crafted invitation has been proven to boost response rates. And response rate is a key concern when conducting surveys, for reasons clearly articulated in this quote from the American Association of Public Opinion Research:

“A low cooperation or response rate does more damage in rendering a survey’s results questionable than a small sample, because there may be no valid way scientifically of inferring the characteristics of the population represented by the non-respondents.” (AAPOR, Best Practices for Research)

With response rate at stake, we need to pay attention to how we write and send out our invitation emails.

This blog post features my most-used tips for writing invitation emails, all of which are included in Internet, Phone, Mail, and Mixed-Mode Surveys: The Tailored Design Method by Dillman, Smyth, and Christian (2014). Now in its fourth edition, this book is the go-to resource for how to conduct all aspects of the survey process. It is evidence-based, drawing on an extensive body of research literature on survey practice.

Plan for Multiple Contacts

Don’t think “invitation email.”  Think “communication plan,” because Dillman et al. emphasized a need for multiple contacts with participants to elicit good response rates. The book outlines various mailing schedules, but you should plan for a minimum of four contacts:

  • A preliminary email message to let your participants know you will be sending them a questionnaire. (Do not include the questionnaire link)
  • An invitation email with a link to your questionnaire (2-3 days after preliminary letter)
  • A reminder notice, preferably only to those who have not responded (one week after the invitation email)
  • A final reminder notice, also specifically to those who have not responded (one week after the first reminder).

 Tell Them Why Their Feedback Matters

Emphasize how the participants’ feedback will help your organization improve services or programs. This simple request appeals to a common desire among humans to help others. If applicable, emphasize that you need their advice specifically because of their special experience or expertise. It is best to use mail merge to personalize your email messages, so that each participant is personally invited by name to submit their feedback.

If you are contacting people who have a relationships with your organization, such as your library users or members of your organization, play up that relationship. Also, make a commitment to share results with them at a later date. (And be sure to keep that commitment.)

Make Sure They Know Who’s Asking

With phishing and email scams abounding, people are leery about clicking on URLs if an email message seems “off” in any way. Make sure they know they can trust your invitation email and survey link. Take opportunities to publicize your institutional affiliation. Incorporate logos or letterhead into your emails, when possible.

Provide names, email addresses and phone numbers of one or two members of your evaluation team, so participants know who to contact with questions or to authenticate the source of the email request. You may never get a call, but they will feel better about answering questions if you give them convenient access to a member of the project team.

It is also helpful to get a public endorsement of your survey project from someone who is known and trusted by your participants.  You can ask someone influential in your organization to send out your preliminary letter on your behalf. Also you or your champion can publicize your project over social media channels or through organizational newsletters or blogs.

And How You Will Protect Their Information

Be explicit about who will have access to individual-level data and will know how they answered specific questions. Be sure you know the difference between anonymity (where no one knows what any given participant specifically said) and confidentiality (where identifiable comments are seen by a few specific people). You can also let them know how you will protect their identity, but don’t go overboard. Long explanations also can cast doubt on the trustworthiness of your invitation.

Provide Status Updates

While this may seem “so high school,” most of us want to act in a manner consistent with our peer group. So if you casually mention in reminder emails that you are getting great feedback from other respondents, you may motivate the late responders who want to match the behavior of their peers.

Gifts Work Better Than Promises

The research consistently shows that sending a small gift to everyone, with your preliminary or invitation letter, is more effective than promising an incentive to those who complete your questionnaire. If you are bothered by the thought of rewarding those who may never follow through, keep in mind that small tokens (worth $2-3) sent to all participants is the most cost effective practice involving incentives. More expensive gifts are generally no more influential than small gifts when it comes to response rates. Also, cash works better than gift cards or other nonmonetary incentives, even if the cash is of less value.

Beyond Invitation Letters

The emails in your survey projects are good tools for enhancing response rate, but questionnaire design also matters. Visual layout, item order, and wording also influence response rate. While questionnaire design is beyond the scope of today’s post, I recommend The Tailored Design Method to anyone who plans to conduct survey-based evaluation in the near future. The complete source is provided below.

Source: Dillman DA, Smyth JD, and Christian LM. Internet, Phone, Mail, and Mixed-Mode Surveys: The Tailored Design Method, 4th edition. Hoboken, NJ: Wiley; 2014.

 

 

 

Last updated on Monday, June 27, 2016

Funded by the National Library of Medicine under Contract No. UG4LM012343 with the University of Washington.