Creating photorealistic stimuli with Blender

My research on the ways that consumers respond to advertising and product packaging is based on empirical studies that use visual stimuli. Generally speaking, the more photorealistic the stimuli are, the better. In addition to Photoshop, I’ve started creating my stimuli using Blender. The results are much more photorealistic than I can accomplish with Photoshop. I mean, Photoshop is a powerful tool, but Blender is absolutely amazing!

As an example, I have made primitive product packages that can be stretched and adjusted in Blender. They are ready for new labels that are created in Photoshop. These models may be helpful to other people, so I will post some of them below.

(1) This image was created by applying this blank label to this 3D model I made in Blender.

Photoshop Contests

Occasionally, I enter photoshop contests. In these contests, people are given an original photo and they are asked to manipulate the photo in whatever way they want. Entries are voted on by people who view the online contest. I’ve won a few of the contests, but mostly I enter them for fun.

I confess that I don’t do these contests very often. It has taken me years to accumulate these contest entries. They’re good for keeping my photoshopping skills sharp. I’m consistently amazed at what can be accomplished with photoshop. I’ve even used photoshopping to create stimuli for my research projects.

I’m sure that Adobe, in the hope of avoiding brand genericide, wants everyone to call them “image manipulation” contests, but most everyone calls them photoshop contests without regard to the software they used. Sorry Adobe. Strictly speaking, I don’t even use Adobe PhotoShop. I’m in the small minority of people who prefer Corel Photo-Paint over Adobe PhotoShop.

– Eric DeRosia

The CSS I add to a new Qualtrics survey

To my eye, the “page width” used by Qualtrics with most of their themes is too narrow. Even on a small-screen laptop, it leaves a lot of side margin.  The issue isn’t merely aesthetic; a wider screen gives me a little more room for wider labels above closed-ended response options.

I use the following CSS code to give a little more width to the space Qualtrics uses for survey questions. This code is added in Qualtrics “Look & Feel,” under “Style.”

Continue reading “The CSS I add to a new Qualtrics survey”

CastingWords: A Case of “Best Practices” with Crowdsourcing

The other day, I had two hours of one-on-one interviews I needed transcribed, and I hired CastingWords to do the task. Before I was a client, I only knew the broad outlines of how they use crowdsourcing (with Mechanical Turk as their broker between themselves and workers).  Once I became a client, I was curious, so I really dug into their process. I even did a few of their HITs as a worker to get a feeling for how they do things. I was really impressed by the way they partitioned the task into atomistic bits. I was also impressed by the clarity of their instructions and by their process of escalating levels of verification and review. Furthermore, I was impressed by the overall experience as a client, including an excellent transcript as the outcome. In terms of process, I submit CastingWords as a “best practices” example of how crowdsourcing can be used.  This guy is correct (a video I commissioned using crowdsourcing with Fiverr as the broker).

– Eric DeRosia

The embedded data setup that I always use with Qualtrics

When I do a Qualtrics q’re, this is the embedded data setup I always use. This setup …

  • imports the participant’s country (as inferred from the IP address)
  • writes the start date and time in a format that’s easier to deal with than the default format
  • creates a variable in the data called “on_mobile” that records “Yes” if the participant is using a mobile device (as inferred by Qualtrics)
  • writes the MTurk Worker ID. This one is unique to my setup, but everything else here will work for anyone.

Continue reading “The embedded data setup that I always use with Qualtrics”

“Women in Ads” Active Learning Exercise

In Fall 2000, I was a young Ph.D. student teaching my second Marketing Principles class at the University of Michigan.  It was in that class that I created my first active learning exercise.  The “active learning” experience was so fascinating for me and so impactful for my students that I began to frequently incorporate active learning exercises into my traditional lectures.  I promised my students that semester that I would post the results of the exercise on the web, and they are still available here.

The learning objectives of the exercise were related to the social impact of advertising.  Advertising is often described (e.g., here) as misogynistic, with images that denigrate and mock women.  Denigrating women is undoubtedly problematic for an enlightened society. The question at hand is whether advertising, collectively as an industry, is at fault.  I wanted my students to directly consider whether this criticism of advertising is accurate.   Continue reading ““Women in Ads” Active Learning Exercise”