Category: ETEC 540

Everyone comes with their own biases and algorithms may be formed with the best intentions, but they are undoubtedly embedded with the bias of their creators. The biases I had coming into the detain/release module were

1. I’ve watched every season of the Good Wife. I think in the last season the main character in bail court.
2. I’ve watched most, if not all of John Oliver’s shows on prison and detainments.
3. The song “What’s Your Story?” came to mind while I was listening to the different podcasts.

I don’t think this background makes me an expert, but this viewing history has made me more aware of biases, an unjust and prejudiced judicial and policing system where sometimes evidence is “lost” or “found” to get a case off the docket.

The charge is serious and sensitive. The information from the algorithm is confusing and makes me think of To Kill a Mockingbird where Tom Robinson was wrongly accused of rape. At the same time, there are plenty of reports in the news where rape victims’ claims were dismissed for no reason. I’d want to see the evidence they have against Alexander Dix. Why is he likely to appear and unlikely to be violent, but likely to commit a crime? Wouldn’t a person likely to appear also be unlikely to commit a crime? Did the algorithm predict he’d commit a crime because of skin colour? What kind of crime does the algorithm predict he’d commit? Shoplifting? Shoplifting is not something I condone, but why would he shoplift? It sounds like he has a family to support. If he is no longer there to support his family, isn’t it more likely that the people he left behind will suffer and perhaps resort to petty crime?

I think the problem with the algorithm is it overlooks underlying issues that need to be solved, such as over-policing in certain areas, lack of social support systems such as welfare and childcare, which O’Neil (2017) noted are related and results in the poor being caught up in “digital dragnets”.

It’s not just adults who can become entangled by these algorithms. LMS, which are becoming a part of more educational systems, can collect data from students that can be used to identify at-risk students. As a teacher I can use the data to confirm what I already know or look into cases that surprise me–perhaps a student who did well in class did poorly on a test due to a stomachache or an argument; however, if this data were sent to the head of school or the school board, the data would not show the wider picture that the classroom teacher is privy to. A report sent without teacher input could unnecessarily alarm and upset parents and students.

Algorithms do have their time and place. They definitely would have made my life easier back when I had 200 students during online classes, but I would have had to follow up with the students’ homeroom teachers to effectively use that data to evaluate and support students.

References

O’Neil, C. (2017, April 6). Justice in the age of big data. IDEAS.TED.COM https://ideas.ted.com/justice-in-the-age-of-big-data/

Infuriating was my initial thought; I think my reaction was so extreme because of some things that happened earlier in the day. After completing the “game” I visited the bagaar.be website and read through their job application page and realized what a great ice-breaker this would be at their job interviews. Now I think the people behind this idea are creative and could be fun in small doses.

To start the game I had to read the text under the giant NO button. While HERE was capitalized, click was underlined, making it look like a link. I clicked on click first because I wanted to check to see if it was a link. It wasn’t a link so I had to click on HERE. Often websites will make the surrounding words a link as well to make it easier for users and in consideration of any missed clicks that can happen. This process to start the activity reminded me of the attention and time needed to subscribe and unsubscribe for paid subscriptions.

A pop-up reminding me time was passing took several tries to close. The option to close the pop-up was written in a small font in a light colour and it could be easily missed when scanning because it looks like a copyright line. I thought this pop-up design effectively took into account the Steve Krug’s psychological insight shared on Brignull’s (2011) article, “[w]e don’t read pages. We scan them”. The time pressure can make people skip or rush through the fine print and click on buttons without knowing what they’re agreeing to.

Another psychological insight Brignull (2011) mentioned is by Robert Cialdini, “[p]eople will do things that they see other people are doing”. Normally I would not bother completing a “game” like this. Even though the assignment says that the “game” does not have to be completed if an explanation is given, I chose to complete the activity. I could have written about my own experiences with dark patterns, but I thought that most of my classmates would complete the “game”.

Before completing this task and after reading about forced continuity practices (Brignull, 2011), I remembered a subscription that I should cancel. To get my cancellation processed, I was asked to confirm my cancellation. The Yes option to cancel was in red on the left and the No option to stop the cancellation was in green on the right; however, I’m not sure if I should describe this company as deceptive because they did send me a reminder email a few days before my free trial ended. They could have set up the options that way to avoid accidental cancellations, but resubscribing should be easy.

Prior to reading Brignull’s 2011 article, I saw dark pattern practices as part of everyday life. I think most organizations try to mislead customers, just like stores hike up prices before sales or make cheaper quality clothes to add to their sales and outlet stores. Just because these practices are common, it doesn’t mean that they’re acceptable.

Dark patterns can go beyond generating revenue; they can be used to drown out honest voices. I was looking for free AR apps to use in an educational setting and it seemed that the completely free AR apps recommended by educational blogs and educators online were out of business. Having never used them, I can only speculate what happened. Perhaps they used honest design in their UI which resulted in lower revenue and less advertising. If honest UI design leads to not having a voice, should honest companies resort to dark patterns to balance the field? Brignull’s (2011) Expedia.com example shows that companies can drop dark patterns, but would they have reached the success they experienced if they had started with honest practices?

References

Brignull, H. (2011). Dark Patterns: Deception vs. Honesty in UI Design. Interaction Design, Usability, 338.

For this assignment, I preferred using the tables generated by Palladio file. First I found the community I had been sorted into on the Palladio file. Grant, Emily, Elizabeth and I shared four common tracks, which have been highlighted in yellow.

I made this table to help me better visualize and make sense of our choices. Both the least and most selected tracks were picked by everyone within my community: The Well-Tempered Clavier and Flowing Streams.

Due to the popularity of Flowing Streams, I wondered if diversity and representation were common themes among everyone’s lists. I read the Task 8 assignments from the people in my community and some random ones from neighbouring communities, and it seems that many people strived for representation of diverse cultures and identities.

During Task 8, I tried to choose tracks with a diversity of cultures and identities in mind, but I was concerned about how I went about this because

1. Diverse representation could imply to extra-terrestrials that diversity is widely valued and exists peacefully.
2. Actually we live in a world where biases are perpetuated in algorithms which allows individuals to live with little exposure to different perspectives (Edwards, 2019).
3. Choosing tracks from different cultures felt like tokenism since it does not address the underlying issues that exist.
4. Representation by population ignores the minority groups that do not have the population numbers or cultural clout to be represented.
5. The data input into algorithms are used to identify current and future trends. How can we be sure that the algorithm will interpret this data so it values diversity?
6. The impossibility of representing all cultures through 10 tracks makes me wonder if a better  way to represent all humans is being overlooked by the majority, thus ignored by the algorithms. The most advanced algorithms use neural networks that are designed copy the way the human mind works and identify underlying relationships within a data set (Edwards, 2019), which means diversity of thought will be limited by the data set.

I looked at Task 8 completed by people in the community farthest from mine. One thing I noticed was the people in this group all mentioned auditory aesthetics, which I feel, were not so prominent in the Task 8 assignments from people in my community. Next I visited the Task 8 assignment of two people who had not chosen Flowing Streams, Johanna and Danya. Johanna listened to the tracks and used universal human experiences, feelings, to curate her tracks. Danya’s process involved using research and sound to pick tracks that represent the world through history, geographical representation, variety of musical instruments, rhythms and melodies to narrow down her initial choices. Like Johanna, she selected her final ten through listening. A completely different approach than mine, but nevertheless effective. Because this data was not in my immediate community, I doubt I would have seen this in any of my searches or social media feeds. If I were to complete Task 8 again, I would seriously consider approaching the task from human emotions because feelings are something everyone experiences and can relate to. Perhaps it would be easier to achieve diversity through universal feelings?

References

Edwards, J. (2019, August 19). What is predictive analytics? Transforming data into future insights. CIO. https://www.cio.com/article/3273114/what-is-predictive-analytics-transforming-data-into-future-insights.html