Thursday, December 04, 2014

#OEB14 - Does Data Corrupt Education

Summary of a debate at Online Educa Berlin. Note that the debate format is not serious and that participants to not necessarily agree with the points that they are making. I am summarizing talks, so if the first person is used, it is the voice of the speaker. Enjoy.

This house believes that data is corrupting education

For: Ellen Wagner

- Data as the meme
- It bothers me that I leave this trail of data everywhere we do
- Data without context is really without value
- data with context is information, and information is power, and power corrupts...
- data in the wrong hands might be misused
- will this change what we do? Absolutely - how many of you as instructional designers like the idea of your profession being reduced to algorithms?
- How do your students feel abiut having every move onserved... cinstrained, analyzed, measured by the data we leave in out tracks
- Taxi apps - Uber - the drivers rate their passengers - if you have been a bad passenger you will not get a ride
- I worry about naive or nefarious uses of data to restrict access or to punish

Against: Viktor Mayer-Schönberger,

- Why does it need the side of attacking this motion in the first place? Seriously?
- Since the beginning of time we humans have tried to make sense of the world around us by observing it, by gathering data
- We already see that learning and looking at data are entwined
- Denying it would not only be foolish, it would be dehumanizing
- our progress of making sense of the world by using data has been nothing short of staggering
- not just knowing more, knowing more accurately
- it would be ridiculous not to use what powers what we learn to inform how to learn
- we are not yet there, we have to move beyond PISA, which measures only outcomes and not process
- the other side wants to make you believe that ignorance is bliss

For: Inge de Waard

- Now what a sly fox you are... using words and making them use different things
- from a digital divide and a data divide it is quite dangerous
- as Facebook links up with big universities, little universities do not get that opportunity
- it costs a lot to access this data - if you don't have the data you can't do the analytics
- you might say, data is open to all - but lets point to Victor, associated with Harvard and Oxford
- and then there are algorithms - if you see a vulnerable group is less successful, you start thinking, let's not recruit from thst group
- 84 percent of government and corporate employees say education is not preparing people for the future - but big data would fix this
- but then I saw a study saying 1 in 3 jobs will disappear, not be changed, just gone - how does data fix this? we have to think about using data for different purposes
- at a time when education is being cut, big data is getting much more funding than ever
- point 2: the norm - a human can't be more than human
- a system represents the thoughts and values of those who create it
- eg. an app analyzes music and determines whether it has hit potential
- but... the masses don't move - the individuals - Gandhi, Jobs - change - if I see them I see they will change the world
- so what is the algorithm of these people? It is not big data - it is very teeny weeny small data that makes the difference
- life is quality, creativity, really living the dream
- data should be shifted toward a new goal, where everybody is included, where we have marginality and exceptions to the rule

George Siemens

- Our two opponents ended siding with us - and they started by attacking my friend of the last two hours
- Why data is not corrupting data: massification, cognitive extension, quality of learning
- in history, the quality of your education was determined by class
- today, the classroom helps, but the biggest barrier to education is poverty
- my opponents want to subject your children to a lifetime of poverty
- we will only massify education through an effective use of data. That's point 1
- analytics as a cognitive extension: like the plough, we use tech to become more productive
- if we were still planting by hand we could support 7 billion people
- in 1550 there was already an overabundance of ideas, so we created classification systems, encyclopedias, etc
- we have always used methods to make data more manageable
- that's what we are doing with data
- through the effective use of data we can manage what was unmanageable. That's point 2.
- Elon Musk, Bill Gates, etc., say 50% of jobs will be taken over by robots
- so we all have to reskill. That will be supported by analytics.
- the assessment process - most learning form you today does not happen in a classroom
- analytics will take those experiences, match them to curricula, and give you credit for what you have done. That's point 3.

Each speaker is separated with a blank line.
- talking about humanity making sense of the world, the starting points is ourselves
- I think simplicity is key, I think data can confuse and blind

- to women, arguing to keep everything the same, two men acting like the hunters...
- Einstein would not agree with you, not all technology is good

- whichever side you think you belong on, big data is coming
- so it's not whether we like it, but how we will adapt to it

- if we accept that the outcome of education is knowledge, and if knowledge is the outcome of relevant information,and if we accept that information is an aggregate of data, then why would we think that data corrupts education

- corruption - almost every other session has been about disruption, moocs, etc...
- why on earth would we not want to disrupt - and corrupt is not such a different word
- why wouldn't we want to corrupt the existing educational systems? (Gilly Salmon)

- the data shows half the world is not connected to the internet
- so how do you collect data from them? Or do you just ignore them?


(George) That's an excellent point - we create tech that creates problems that can only be solved with more technology. Even in low-data environments, you're already collecting some data. We can use mobile data, human observation, etc.

(Ellen) All data are not created equal. We are equating different types of data. Until we are better at understanding what we are talking about we should maybe be careful. Remember it took 30 years before people thought online ed was a good idea.


Data is corrupting education, and I hope we are riddled with it soon. It is the natural order for it to corrupt education. The system is broken, it was built broken, so wither stay on the sinking ship or hope for something else.


(Ellen) I like the idea of disruption, but 'corruption' is a self-serving deviation from an ideaal. So we have to be careful here.


- if we are talking about big data, we are talking about access to data, analysis of data online
- but it's not data we want, when I see my students, I don;t want to see data


(Viktor) the motion concerns data, not big data. I am not here to defend data. But I have children and sometimes care about them too much. What we want is for teachers to focus on what they do best. So we also need to help them get better, and the only way to get better is to get feedback, and the only way to get feedback is with data. I depend on data.

(Inge) I want to add that data is of course positive, but there is an example of data that has not made any data though we know its effects: climate change. We have data but we don't act on it.


- key words - privacy, ownership, corruption, Snowden....

- if something that corrupts education is something that stops it happening, then it's not data that corrupts education, but it's subject specialists
- when the data becomes the priority but it's not used meaningfully, then it becomes corrupting because it shifts attention away from teaching

- the words education and data have not been defined
- education, if we talk about the big structure of education, then it can be corrupted
- if we change from 'big data' to 'open data' then we don;t have a problem, but then owning data creates power, and that power is corruption

- at the management board of the university we are glad to have data because professors would never change otherwise
- it's not big data, it's small data, just a few students unhappy with the way professors are teaching

- I want to compare with farming - we had to mechanize
- probably we will discover many efficient ways to improve efficiency

- it's not whether data will corrupt education - we agree it will - it's how we should use it

- something you haven't pointed out, is that data can be corrupt
- if data is corrupt, what does that do to education?

- I think the rwal danger is that we're talking about beliefs
- I think we should have facts, perhaps a large survey, so we can have a proper debate

Against: Viktor Mayer-Schönberger

- for Ellen and Inge it was a significant challenge, and they rose to the challenge
- we had the easier task, data doesn't disrupt education, that's clear
- we talked to Sal Khan who told a story about a small child in California who just didn't get math, didn't get it through the videos, data was collected, and they readjusted the videos, and suddenly she got it, and by the end of the summer she was the best in her math class
- it's not about the content, it's about the data we can collect to learn how to improve learning

For: Ellen Wagner

- taking a position in support of the motion is not such a stretch
- my actual experience with data that is corrupt, that has not been cleaned, that is in the wrong hands, makes this clear
- I wonder whether it's the data we should be worried about, but whether it's the people
- there is no algorithm that defines what is a good person
- we need to have more than data helping us make decisions


Ironically, we had a vote

from app: 27% for, 72% against
in room, same percentages


Your comments will be moderated. Sorry, but it's not a nice world out there.