Subscribe
Share
Share
Embed
Join feminist coaches Taina Brown and Becky Mollenkamp for casual (and often deep) conversations about business, current events, politics, pop culture, and more. We’re not perfect activists or allies! These are our real-time, messy feminist perspectives on the world around us.
This podcast is for you if you find yourself asking questions like:
• Why is feminism important today?
• What is intersectional feminism?
• Can capitalism be ethical?
• What does liberation mean?
• Equity vs. equality — what's the difference and why does it matter?
• What does a Trump victory mean for my life?
• What is mutual aid?
• How do we engage in collective action?
• Can I find safety in community?
• What's a feminist approach to ... ?
• What's the feminist perspective on ...?
Becky Mollenkamp (00:01)
Welcome to Messy Liberation. This is our special Everything is Political series. Where we're going to help you draw the connections between your everyday life and the world around you. Let's get started.
Taina Brown she/hers (00:14)
we have Millie with us today for the part of our Everything is Political series. She and I have known each other for, how long has it been now? Four or five years? Six maybe?
Millie (00:22)
Yeah, it's been a while
since like early pandemic days.
Taina Brown she/hers (00:25)
Yeah,
yeah, I think we like connected online as that time went with a lot of people. And so then you work with data, And that statistics and all that. And so I thought this would be an interesting conversation to have for the series. So tell us a little bit about yourself before we like jump into the conversation.
Millie (00:28)
Yeah.
Mm-hmm, yeah. So I would say I'm a woman of many things, but in my day-to-day role, I do data and analytics. I've been in this space for, I want to say, pushing a decade now, which is crazy. I started off in the higher education space, doing research and program evaluation for large city-wide initiatives and moved into the private sector around the... ⁓
pandemic, continuing with a little bit of education, but then jumping into the e-commerce space, doing analytics engineering type work, and now in the tech space, continuing in the business intelligence space. Using the same kind of set of skills in different industries and domains. Otherwise, I'm just another girlie who loves to walk in the park and look at turtles.
Taina Brown she/hers (01:33)
Turtles, yes.
Millie (01:34)
and dance
on occasions. Yeah, that's a little bit about me.
Taina Brown she/hers (01:38)
Excellent, excellent. So I'm going to start us off with the question that we've been asking everyone that we've had on the series. that's, at what point did you realize that the space that you working in was political in terms of like, understanding how power can fluctuate, how power can become dynamic in the work that you do, depending on right perspective and bias and, and what agendas, know, or intentions or strategies might be present.
Millie (02:04)
Yeah, I would say like pretty much immediately. Like as soon as the thought of the idea of me going into like data as my career, because I mean I thought that I was going to do maybe like some clinical work like cyber psychology or like because I did my bachelor's in social. I quickly realized like no I care about like big problems and I do like
Taina Brown she/hers (02:07)
Yeah.
Millie (02:29)
the research aspect of things. So from my own studies, I understood that the world and everything that is important to me is seemingly political. And then for the things that I wanted to do, which was in the education space at first, that is also very political because that's very much in and often used as a carrot for communities to have access to resources or even
have a level playing field. So, and the way in which data is available or not in that space or any space, like I immediately was just like, all right, this is a thing.
Becky Mollenkamp (she/they) (03:05)
Right. This is a thing.
And when we were talking about this, I was like, Taina, you have to take the lead, because I don't know much about data, which actually is why I want to start here, which is for all the people are like data. First of all, if you're like me, you're like you hear numbers and math and your eyes start to glaze over. It sounds dreadfully boring. Right. Like exactly. For a lot of others, like, don't care, not interested. And yet even as somebody who fits in that squarely inside of that camp.
Taina Brown she/hers (03:22)
out of the ears. ⁓
Becky Mollenkamp (she/they) (03:32)
I can also realize that data affects everything, because even with this podcast, we're looking at analytics. I examine to see who's listening, how many people are listening, what topics of interest are not of interest, all that kind of stuff. I think obviously they're using the census as something we think of as a giant data set that is really important in affecting public policy. know health care around studies and who does and doesn't get studied. All these things actually matter, and data is like in all parts of our life.
Millie (03:39)
Mm-hmm.
Becky Mollenkamp (she/they) (04:01)
So I guess just why is data, like what you do, analytics data, analyzing numbers, thinking about research and big sets of numbers, why does that matter to people? For those people who are like, it's just numbers who cares. Like, why is that something that is important? Because you said you care about big important things. And I think some people might say like, well, it's, yeah, that doesn't sound, it doesn't sound like the person who's, I don't know, writing the policy or whatever the thing might be.
Taina Brown she/hers (04:22)
with just information.
Millie (04:24)
you
Becky Mollenkamp (she/they) (04:28)
What is it about data that appeals to you and why is it something that really does matter in the big picture?
Millie (04:34)
Mm-hmm. Yeah, mean like because data data is information. It's and I am just the kind of person that just like likes to know things And like though even with like when I'm trying to decide what I'm going to buy like looking for an umbrella like you would think would be like a simple thing but like I'm going to I'm trying to find like what is the most efficient umbrella, but that is cost-effective You know and like all
information that you can gather or is out there for you is all data. And I think the way in which your world is shaped, especially today, with the way in which the internet is vastly different than where it started to where it is now. Whereas before, it was kind of like a Wild West of a place.
Taina Brown she/hers (05:15)
Mm-hmm. Mm-hmm.
Millie (05:16)
Everything was kind of like equal. Everyone had like a similar experience when they opened up a browser and they typed something, they will all get the same results. Now you get very tailored results because it's based off of the data about you from like your actual computer to things that you typed in before. Yeah. All of the algorithm like results of all people, like what they respond to or feel is relevant. That's all data that has like been given. so
Taina Brown she/hers (05:32)
your algorithm.
Millie (05:44)
So yeah, if you are not paying attention to those things or if your representatives in government aren't considering all those things, these are all the things that impact your life. So I think in that sense, from the small things to the big census data, these are the things that you want to pay attention to for sure.
Taina Brown she/hers (06:01)
Mm-hmm.
Yeah, yeah. That's interesting that you
said, you brought up the example, like, okay, if I'm buying an umbrella, because it made me automatically think like, whenever I am making a purchase, or even like looking up trying to find like a restaurant, right, to eat, which this is a point of contention between me and Mel all the time, like, what do you want to eat? I don't know. Well, do you want this? I don't know. Well, do you want this? I don't know. Right. And so,
Millie (06:17)
and
Becky Mollenkamp (she/they) (06:26)
I'm convinced that's the number one cause of divorces by the way. It's gotta be. And breakups.
Taina Brown she/hers (06:28)
It's just not knowing what you want for data, right?
Millie (06:28)
Hahaha
Taina Brown she/hers (06:31)
But like, I never thought about the way that I researched those things in terms of looking at data, right? Because what my normal protocol is for figuring out what restaurant to pick or what brand of umbrella versus another brand of umbrella to pick is I look at
the highest rated reviews versus how many of those reviews are present, right? So I'm like cross matching data, right? Like I'm cross referencing data. So if it has, if it's a five star umbrella, but it only has 10 reviews, but then there's like a 4.2 star umbrella, but it has like 2000 reviews, then that seems more, like a better purchase to me because more people have like engaged with that product and have like reviewed it.
et cetera, same with restaurants. But the other thing is, OK, so that's on a really individual scale. But it also made me think, so I do this Black History Month presentations sometimes with organizations and stuff. And one of the things that we talk about is, or that I try to bring up, is the importance of accurate data that drives policymaking when we're talking about
especially things like the census, right? And so, and one of the examples that I like to use is like the organization Lean In, right? They have this data about the wage gap, right? But in their statistics about the wage gap, the way that they have women divided up, they have both ethnicity and race on the same like,
level playing field, if you know what I like on the same graph. And those are two different things, right? So like, for instance, they'll have like Hispanic women and then like Black women, but like, you can be Hispanic and Black, right? Because one is an ethnicity and one is a race, right? And so when you're looking at data like that, like, and that is what's driving policy and you don't have accurate representation or accurate numbers, then the policy is not really going to like hit the mark completely, right? Like it's going to be skewed from the get-go.
Millie (08:14)
Mm-hmm.
Taina Brown she/hers (08:39)
And I think a lot of people, when we think about just information, like why is it important? what is the need for everything to be so precise? A lot of people just assume that like data is something neutral. And this is something that Becky and I talked about before in terms of like when we started talking about having you on, right? It's like people just assume that data is like a neutral subject. Like when they see statistics on the news, that it's neutral, right? That it hasn't, there's no bias in that.
Millie (09:01)
Mm-hmm.
Taina Brown she/hers (09:06)
that it's just numbers, it's just figures, it's just a graph. But that's not always the case, right? Can you talk a little bit about that, like how data can be manipulated to not be neutral to get a certain result from either people in the company, people in the public, et cetera.
Millie (09:06)
Yeah.
Mm-hmm.
Yeah, yeah, no, that's a good point. I think like bass level there there's like
a general understanding that data has bias injected into it because humans are biased in general. That's like our natural state. We cling to what is familiar or what we understand. And so that's going to be reflected in the data and the way that we collect it. But then even with that, there are methods statistically to try to help combat that. ⁓
Taina Brown she/hers (09:52)
Mm-hmm.
Millie (09:53)
even when you're doing survey data. even with that, yes, there are instances where people will...
use numbers in a way to influence their argument and maybe leave out parts of another just for the sake of showing something. One of the things that's as you were talking that came to my mind, I was just trying to think of a big example of where, and this is maybe not fully related to manipulation of data per se, but.
I think it gets more towards the bias of it all. So for example, polling data when it comes to elections.
Taina Brown she/hers (10:27)
Mm.
Millie (10:28)
Polling data is already biased, and they'll random sample people from all over so that you have less of that, or you have more of a representation of all kinds of people of all different thoughts, putting in responses to who they plan to vote for. But for the 2016 election, polling data for the longest was pretty reliable of just who would come out and win. But that year, given that particular race,
was so unlike any other race before. There was nothing in the data that would help show that it would be remotely possible to have the outcome that we didn't end up having. And I think from there people were just like, yeah, you see, the data is always wrong, you can't believe these kinds of things. it's just like, there were a lot of things that were very different than the past that you just can't get data to do or show.
But even for news articles, for example, one of the things that you'll see, depending on the source, but for example, sometimes I even have commented on the New York Times and the way that they do infographics, because you want things to be understood for most people, and you don't want that to come at the cost of, you don't want to do that at the cost of aesthetics, but at the same time it needs to be understood. But you'll see things like,
Taina Brown she/hers (11:33)
Mm-hmm.
Millie (11:43)
⁓ changing the access, like the access to like the numbers on the line and like not making it start at zero so that the shape looks a little bit more dramatic than it actually needs to be. ⁓ Or they'll like pull out a line from a research study. And this is what a lot of articles do actually when they are quoting like
Taina Brown she/hers (11:53)
Uhhh
Millie (12:04)
larger research studies or peer-reviewed papers that, you know, they have a whole methodology and lots of limitations, but we're not gonna quote all of that in the article. We're gonna pull out the one or two lines, but we're not gonna say that this was based off of a small clinical study of 50 people from Minnesota at this time. We're going to, they'll blankly say it, and people are not gonna click and go through and read it for themselves.
Taina Brown she/hers (12:26)
Mm-hmm.
Becky Mollenkamp (she/they) (12:30)
I am like go.
Millie (12:34)
moments
like that is where I feel like those examples of like those biases or like the manipulation of data, whether or not it was like intended, it often happens, yeah.
Taina Brown she/hers (12:44)
Yeah, yeah,
yeah, yeah, just like, like, it makes it easier for people to take things out of context, right? And so like, I know usually, God, being in fucking grad school, just like, you can't enjoy anything, I think, once you go through like, like any kind of really like critical thinking kind of session with a professor or whatever, or in school or whatever, because like, anytime I like look at like a study or something, I'm like, well, how big was the sample size?
Millie (12:50)
Yeah.
Mm-hmm.
Taina Brown she/hers (13:11)
Who was
doing the polling? Like how were the questions phrased? Right? Because even something like, I mean, my wife works in data, right? Like this has come up in the podcast before, but like even like watching her try to phrase questions for surveys in a way that isn't going to like be like a leading question to a very specific outcome and try to keep the question or the phrasing as neutral as possible, knowing that it's not going to be completely objective. Right? And I think
And I think this is where people who are sometimes still like in this world are a little bit delulu where they have this like idea that like objectivity is like complete. it's like a, like you can be completely objective when you're like doing research, which is not true because like, as you said earlier, like we all have biases, right? And I think people hear that word, well, I'm not biased.
Millie (13:47)
you
Taina Brown she/hers (14:05)
Well, yeah, we are. That's necessarily a bad thing. That just means you have a perspective. You have experiences that shape how you see things. And so that is going to intentionally or unintentionally become embedded in the work that you do and how you consider what questions to ask, how to ask those questions, how to figure out where to get people to answer the questions that you're asking. ⁓
Becky Mollenkamp (she/they) (14:09)
Thank you.
Millie (14:31)
Mm-hmm.
Taina Brown she/hers (14:31)
But then I've also seen like, where sometimes I'll see like a graph or something and it just, it looks so messy and so confusing that it's like, I don't even know how to begin to understand this. how, like who put this together?
Millie (14:49)
Mm-hmm.
Taina Brown she/hers (14:50)
Who is this? Who is the intended audience for this? Because this makes no sense to me as someone who's not in data. Like, I couldn't even begin to tell you what this information is telling me. Yeah, it's just, it gets really meppy.
Millie (14:59)
Yeah,
even like the use of color too for like graphs, like people will intentionally use specific colors to like highlight things, which is a good practice for like a data visualization. like, yeah, using red is certainly a choice of just like alerting something that may be not as like something that should be alerted or something as like, you know, key. Yeah. Or as you were talking to, I was thinking like even something as benign as like
Taina Brown she/hers (15:04)
Mmm, mm-hmm.
Yeah.
Becky Mollenkamp (she/they) (15:23)
As you were talking.
Millie (15:26)
uh like family feud. I don't know if you like have like come across like those like uh like either on like TikTok or Instagram of the there was like one time where they asked 100 people like the best like hip-hop artists and they were just like they they weren't even like people that you would that most people would necessarily think of but it was like whoever they asked in whatever like you know area where they go on the street and like pull for these people these were the
responses that they came up with and it was I think it was a celebrity one where they were listing out you know like probably like the top Billboard ones of like you're like you're Tupac you know and and they were getting things like Eminem like they weren't even getting like a little weight like it was like all these and they it's
Taina Brown she/hers (16:06)
No
All these like random wannabe rappers.
Becky Mollenkamp (she/they) (16:17)
Yeah, well, and that's
because to your point, Tanya, I think people often will hear things like the numbers don't lie, right? That numbers in facts and figures, those are facts and figures. They're facts. They're not they're not human interpretation. But what does matter, you could say it can be a fact that we interviewed 100 women or let's even say we interviewed a million.
So it's like a huge number, the data set's big and we interviewed a million women and we asked them these questions. And if all of those women were white or all of those women are able-bodied or all of those women share certain characteristics and then you're asking them questions and now saying, because we've had such a large sample size, ask women, we take that and we from that extrapolate out and say, right, well, women feel this way about this thing. That's still not.
Taina Brown she/hers (16:58)
General.
Becky Mollenkamp (she/they) (17:03)
truth, right? Those numbers aren't maybe lying, but they're not sharing the whole truth. some of the, I think part of what I want to get to is why that matters, because it's fine, whatever. Data is used in all sorts of ways all the time. It's never like, we're making it clear. It's not always sharing the full reality or the full intersectionality of what data can give us. But these things have real world consequences too. It's not just like,
Millie (17:11)
Mm-hmm.
Taina Brown she/hers (17:27)
Mm-hmm.
Becky Mollenkamp (she/they) (17:28)
while it's important to have better representation on a game show, that has a limitation on how much it matters, right? But there are really important reasons that manipulating or even if it's not, like you said, intentional and malicious that manipulation kind of makes you think of, even if it's unintentional, know, biases that show up in the numbers, these can have really profound effects on people's direct lives.
Millie (17:33)
it.
Becky Mollenkamp (she/they) (17:54)
I'm wondering if you can share some of ways that you know of that getting the numbers wrong really matter in our day-to-day lives.
Millie (18:02)
Mm-hmm. ⁓
Becky Mollenkamp (she/they) (18:03)
I mean, I immediately
obviously think of things like gerrymandering because that's going on right now, right? But I know that there's plenty of it. So I'm just curious what comes up for you when you think about
Taina Brown she/hers (18:06)
⁓ yeah, yeah.
Millie (18:12)
Yeah, I think with like with like census, for example, having like an accurate understanding of like who is in the population helps with like allocating certain funds to states. And that's why there's like a lot of like grassroots groups that will try their best to like do more like door to door in a representative type of way to have people like open their doors and potentially like, you know, give information.
because of that, it impacts the amount of money your estate could get for certain things. Or even in education, understanding how many...
children there are that have specific needs or who are of low income, if all of that is not accurately shown, the school may not get the amount of money that they're supposed to get from the federal government for those kids to have what it is that they need, which then, of course, impacts the community. schools also often double as community spaces where you have access to library resources, where libraries aren't
there or other social goods that schools will help double as, even sometimes like healthcare and some really ⁓ sparse spaces. Yeah, those are the two things that immediately come to mind that feel pretty prevalent.
Becky Mollenkamp (she/they) (19:26)
think about healthcare
too, because I know data sets are really important in research studies for drugs and medications. And from my understanding that because of racism and sexism, historically, large groups of people were not included in a lot of studies on medications. So dosages would be wrong or they weren't correct. They weren't studying, you know, diseases and other things that might affect certain populations more than others and that kind of stuff as well, I would assume.
Millie (19:34)
Mm-hmm.
Taina Brown she/hers (19:45)
Mm-mm.
Millie (19:53)
Yeah, no, aspirin I feel like was one of them, where it was was white men that were studied in a small group of them, but later on they've learned that the dosage is different for women or female bodies. And that's only up recently, or even studies on the amount of blood absorbed for period products.
did not use for a while, like most of them, think like just used water or didn't like use like a compound to like simulate blood and or blood clots. Yeah, which I'm like, that is wild. What do you mean?
Taina Brown she/hers (20:25)
The viscosity of blood, yeah.
Yeah, it's like it's literally a
containment product and it's not using, you know, something that's accurate to that.
Becky Mollenkamp (she/they) (20:37)
Yeah,
it wasn't Plan B only, I mean, originally anyway, studied and made for people who weighed under a certain amount like 175 pounds.
Taina Brown she/hers (20:44)
Yeah, yeah,
if you you weigh over a certain amount, it's not effective. Like Plan B is not effective for you. And they don't they don't really advertise that. Like that information, like it's out there if you like do your research or if you have a good health care provider. Yeah.
Millie (20:45)
Yes.
Becky Mollenkamp (she/they) (20:48)
Exactly.
or if you happen have a good pharmacist who understands
because then what happens with this data is so much of it is left on the end user or other people who can sort of understand data at the level we're talking about because it's not as simple as just looking at some numbers on a pie chart for the kinds of things that you're talking about here. It takes a pharmacist who actually cares, will read studies, go deep into the understanding of the data to be able to really, or a doctor or whoever.
Millie (21:19)
Thank you.
Taina Brown she/hers (21:20)
Mm-hmm.
Becky Mollenkamp (she/they) (21:23)
be able to educate you on these things and if that's not happening or they never get that education, then mistakes happen and people end up pregnant who don't want to be.
Taina Brown she/hers (21:30)
Yeah.
Millie (21:31)
You
Taina Brown she/hers (21:31)
Yeah,
no, I knew someone who was on birth control and got sick and their doctor also put them on antibiotics. And if you take antibiotics when you're on birth control, it makes the birth control ineffective. And the doctor just didn't put that together. And she ended up like getting pregnant. I also know someone, you know, talking about like policy and like funding and things like that, like who was incarcerated.
a white Latino male. And when he was looking at his record or his paperwork, they had him listed as Caucasian, as white. And he was like, I'm not, that's not my race, that's not my ethnicity. I'm Latino. And one of the officers point blank told him, we do that. So the numbers are skewed. So it doesn't look like.
we're incarcerating as many black and brown people as we actually are. Like literally just like set it outright.
Millie (22:24)
Oh my god, yeah, I forgot about incarceration data. Speaking of going back to the point of data being facts and often used to make a point that isn't always true, that is very much the case with incarceration data because when you look at crime stats in spaces, more often than not, Black and brown people are the ones that are arrested. And then they like to use that, say, oh, well, they're committing more crimes. Therefore, they're arrested.
is obviously very much false because of the bias of people toward black and brown people. And they're more likely to be arrested or not even given other opportunities to deal with whatever the situation may be. Yeah, things like that. Heavily biased data used incorrectly.
Taina Brown she/hers (23:14)
Yeah,
and at one point I think the bias there too was that like black and brown people compared to the percentage of the population that they made up was smaller than like white people, but they were incarcerated far more likely than white people. So it's like, looked like they were committing more crimes, but really they weren't. It's just like compared to the actual population numbers.
Millie (23:27)
Mm-hmm. Mm-hmm.
Mm-hmm.
Taina Brown she/hers (23:38)
It
was like they actually weren't committing as many crimes. It just looked that way because they were only using like a contained data set within that specific population as opposed to comparing it to other populations. so that goes back to your point earlier about like taking data out of context too, right? Without like the whole picture, without the whole context, like it's easy to misconstrue what's really happening and how to like...
Millie (23:51)
Yeah.
Taina Brown she/hers (24:06)
address what's happening, right? Like how to provide interventions or solutions or policies that can help mend whatever is happening in society.
Becky Mollenkamp (she/they) (24:17)
Data is a really powerful way, as so many of the things you're talking about here, to reinforce the status quo, to maintain the systems as they are by manipulating, pulling, misleading about data that then reinforces a lot of policies. I feel like law enforcement, quote unquote, is absolutely a place where we see that happen all the time, where data is getting completely skewed.
to be able to justify and reinforce things that are happening. I think it's happening also, again, because it's sort of part of that carceral system is with the Department of Defense and war and that sort of thing and immigration. But all of that says that there is a lot of like human intervention into the numbers that don't lie, right? So the numbers themselves may not be lying, but all the ways that humans are going about interpreting and presenting the data.
Taina Brown she/hers (24:54)
Mm-hmm.
Millie (25:04)
and then.
Becky Mollenkamp (she/they) (25:10)
can absolutely tell very different stories depending on who's doing that. has now come along, right, as this, with this promise to be able to take out some of that human element. think, you know, I think we all know what that where this is going and how we're going to have this conversation. But I'm curious for the people who might say, well, AI is going to come along and help solve so many of these things because it removes so much of that human element from it. It's using a computer also, which, you know,
Taina Brown she/hers (25:26)
Okay.
Becky Mollenkamp (she/they) (25:37)
doesn't lie to interpret data and present data and stuff. So that's going to help. I can feel that that's probably something that's being said and used in the promotion of AI in your space. But you tell me, what are you hearing? What are the promises being made about AI when it relates to some of this? And what do you think is the reality?
Millie (25:55)
AI, like it's, don't, at least for where I'm at right now, I'm not hearing folks like speak about AI as like removing those biases because one of the terms that like you'll hear all the time is like garbage in, garbage out. So if biased data pulled from the entire internet is going in, we're going to get bias information out. And like,
for the fact of the matter is like if you were to like oversimplify what like AI is doing it's just like doing a really good job at predicting the next word in their phrase right.
So it's not, it doesn't have its own sentiment. It only understands the context that you give it if you give it context. If you don't give it context, it'll go wild with whatever it feels is relevant to the nearest grouping of a relevant word.
But because we are also biased, even with our own context, we're gonna give like biased contents back. So like even like going like for like a technical example.
you like I can have like a conversation with with an AI model and ask it something like you know tell her whatever the situation is and like ask it a specific question and another person can do something similar and like we could get similar answers and just like kind of like any Google search you can get similar answers but also you can get like pretty different answers because it is also structured to be very psychophantic so like it's it's trying work
is trying to please you, not that it has feelings or whatever, but like, know, for the simplicity of it all.
Taina Brown she/hers (27:32)
It
learns you.
Millie (27:33)
Yeah, yeah, it learns about like from everything that you type or what you don't include. And also for these companies who create these chat bots that is the most familiar we are with AI today, they're using models that are successful and the ones that are most successful are the ones that we feel validated from, because we all just wanna be validated. But then that creates even more
biases in specific contexts that are not presenting, it's not necessarily a truth, but it's really good at making it look like it's true because it's perfect and it can do everything really in a way that just seems like it's an authority because it's writing perfect sentences and doing it so quickly, faster than we can even imagine doing it ourselves.
Taina Brown she/hers (28:18)
Mm-hmm.
Millie (28:26)
So yeah, can't, at least for the where I'm at, I'm not seeing this as like remotely helping in that aspect. I do think it has a lot of power to help with doing research and like getting through lots of information a lot quicker and being able to like be more efficient on things. But there's still a long ways to go before we like get anywhere where it's not like.
you know, that it's not like having its own like, internal agenda.
Taina Brown she/hers (28:57)
Yeah,
yeah, yeah. Yeah, I've heard, you know, in just different conversations where, you know, people who do research using AI to like, to help speed up the process of the research, but not to like, do the actual research, right, not to rely on it to like actually interpret the data, but to like speed up the process in which the data gets interpreted, right. And I think, I think that's a helpful way to use it, right.
to be able to get to answers faster, to get to a place where the actual humans who are doing the work can do better work and do not necessarily faster work, more impactful work because they have more time to spend doing their interpretation as opposed to organizing the information. It's such a long process to do data analysis.
Millie (29:38)
Mm-hmm.
Taina Brown she/hers (29:47)
from the little that I know. I can't even imagine how much more work it is. And so I think that's a good way. But I like what you said about garbage and garbage outright, because I think a lot of times the way people talk about AI, it's like this godsend where it's like, well, AI told me this or AI told me that. And it's actually just telling you what somebody else already said. It's not necessarily feeding you new information. It's just organizing the information that's already available in a way that
Millie (29:49)
you
Taina Brown she/hers (30:14)
is in a way to answer the question that you're asking in a way that sounds pleasing to you. Like it's not necessarily innovating, right? It's not creating knowledge or creating information at all. Did you have another question? I want to shift this a little bit, but I didn't know if you had another question while we're on this topic,
Becky Mollenkamp (she/they) (30:29)
Did you?
Well, I do.
I think so related to this, although I want to make sure we get to data as a product as well, because I think that's really important. But how I don't think the answer sounds to me like the answer is not that there will ever necessarily be a way to procure, create and like interpret data in a way that removes all bias. That is just necessarily part of the process. It could be wrong. So tell me if I'm wrong there. But if we're just sort of saying
if humans are involved or even AI built by humans is involved, it's going to have bias baked in that we can't strip that out. If that's the case, then what do for people who are listening, who are less likely to be involved in actually managing data, but more likely to be on the side of I looked at the New York Times and I saw this survey that said, or, you know, whatever that is there, they're the ones who are left sort of responding to data that's being presented to them.
Taina Brown she/hers (31:17)
.
Becky Mollenkamp (she/they) (31:24)
How do we become better consumers of data? If we know all the data probably has biases in it, then how do we as consumers start to believe data, understand data, live with data?
Millie (31:24)
Mm-hmm.
Yeah, I feel like while it is both easier now to have multiple sources and gather that all together, it's harder to make sense of it and trust it. But I think...
I think the old school method of just like pulling in multiple resources and you can use AI to like do that too. Cause like, yeah, if there's something that I don't want to spend all of the time, clicking out all of the links, I can talk to a chatbot of just like, you know, helping me gather all these like resources.
and also asking it to give both sides of an argument and where those sources are coming from and decipher from those are these reputable or not. think...
I think one of the things that is becoming harder today and that needs to come back is a sense of data and media literacy in general. Because I think we've been treating blog posts or short form video the same as a New York Times article or other legacy media who has very different standards and practices.
as all one in the same. And sometimes they'll be referencing those like reputable sources and that's great. And I do love those because we all don't have time for all the information too. Like I love me a like resource to like follow of like a one minute video of someone who has read the things for me and has like kindly put a link to maybe like resources or even like podcasts and whatnot. Like it doesn't have to feel like work.
Like you can pull it from wherever you need to be, I think having a variety of sources to like pull from and using your like judgment based off of like good standards of practice when it comes to like data. like, like for example, like what, what, what you all have like listed out of just like large enough sample size, like, ⁓ you know, understanding like what's the representation there.
how long has this been happening? Like how long ago was this like information poll just didn't need to be updated? As a, yeah, as like your like base level. But again, I think AI can help if you give it the good prompts to go with that.
Taina Brown she/hers (33:54)
Yeah, yeah, yeah.
Yeah, I think what you said about the literacy part of it is so important. even what you said earlier in the conversation about one of the graphs that you saw, it didn't even start at zero, right? To just kind of create a more impactful, a more striking image.
Millie (34:06)
Thanks.
Mm-hmm.
Taina Brown she/hers (34:11)
So just
paying attention to details like that is so important, like knowing that like a graph should start at zero. Like if you're looking, you know, if it's being presented, it should start at zero. shouldn't start at five, it shouldn't start at 10, it shouldn't start at 100, right? I mean, if it does, that should be clearly stated, right? It should be obvious that that's the case. It shouldn't be something that's like obscured or hidden. Like I think one time I...
was watching the news and it was during an election cycle and they had this poll and it was a national, I think it was one of the presidential elections actually, and it was a poll of 2,000 people. And I was like, okay, how accurate is that compared to the entirety of the United States? That's such a small number compared to the millions of people who live in the US. But also, where were they polling people? And I think that's where people just like...
Millie (34:45)
Mm-hmm.
.
Taina Brown she/hers (35:01)
they look at information that's being presented to them and they just like don't ask any questions at all. And I think that's where the literacy part comes in. Cause it's like, okay, well who was being polled? Where were they being polled? What were the demographics of the people who were being polled? Right? Were they like suburbanites? Were they, you know, in the city? Like was it New York? Was it Wisconsin? You know, was it Idaho? Like, cause those, those things matter as Becky alluded to earlier because
Millie (35:06)
Mm-hmm.
Taina Brown she/hers (35:28)
all of those social categorizations, that's what creates the bias that we each have, right? That's what creates our perspective. And so if you're only polling people who live on a potato farm in Idaho, not that there's anything wrong with that, right? People on potato farms in Idaho need representation, right? They need good policy just like the rest of us. But if that's where you're polling and then you're mapping that onto people who live in
Millie (35:44)
Mm-hmm.
Mm-hmm.
Taina Brown she/hers (35:55)
Brooklyn, New York, like there's a mismatch there, right? The needs are different.
Millie (35:59)
Yeah,
well even with that though, I do think that there needs to be a level of trust for those that do do this work and have the credentials to do research or polling like that.
because more often than not, they're like, you know, from like a research like think tank, who's like doing this, and it's like bringing presented on the news that they have like, tried their best to like get a representative sample, and they're using statistical methods to like, make it so that it does make sense to take from that sample and apply it to the larger like population. ⁓ So like, so yeah, so like, so I agree and like, people should then like ask those questions too.
Taina Brown she/hers (36:33)
Hmm. Hmm. Okay.
Millie (36:40)
but also like, you know, take note of the methodology used because we should still trust the experts. Because like I'm also not one who's like out here to be like, don't trust the scientists because like they do know. ⁓
Taina Brown she/hers (36:54)
It's all a
conspiracy. Yeah.
Millie (36:58)
Yeah, let's not go too far with the conspiracy. We should
trust the research thing things. Even if it seems like a small sample, it might be the best that they could at the time, but it still could be representative and balanced in one ways and maybe not so balanced in others. And they will probably note that. But yeah, but if it's a BuzzFeed poll, yeah, OK, that's something else. So we could question all of that.
Taina Brown she/hers (37:19)
Yeah, yeah, yeah.
Becky Mollenkamp (she/they) (37:20)
Thank you.
Taina Brown she/hers (37:21)
it's
like infotainment, right? And I think this is where people don't understand the difference between information and entertainment and infotainment, right? Which is like the budget stuff where it's like, technically it's information, but it's for the purpose of entertainment, right? It's not necessarily for any other real big purpose. What was I gonna say? But yeah, but like, I think where I get skeptical is like, if they don't make it clear, like if they don't share.
Millie (37:23)
Yeah.
Yeah.
Yeah.
Taina Brown she/hers (37:47)
this was our polling sample, this is where we went, like this is what we did. If the methodology, if it seems like they're intentionally obscuring the methodology, then I'm just like, what's happening here, right? Can you really quickly explain what methodology is for people who have never heard that word before, don't know what it means?
Millie (37:55)
Yeah.
Yeah, I would say methodology is like a ⁓ set of like practices, practices or like rules. So for research, if you go back to like grade school, like research thinking of like the scientific method, you start with like your problem statement and then you have a hypothesis and like that hypothesis should be phrased in a way that you can either accept that statement as true or reject it as not true.
with whatever data that you have. And there are all different methods of how to analyze the data that is chosen and that has further details to go into. And with that, all of that should be then given with conclusions and whatever limitations that you have. ⁓ So that is the science of method, which I think is just the blanket other term when it comes to talking about methodology. And there are other options
and nuances of ways in which you can go about studying something, but I would say that's the boiled ground of it. Yeah.
Taina Brown she/hers (39:04)
Yeah, yeah.
is it standard practice? is it, well not standard practice, but is it best practices for people who are doing research to clarify what their methodology is in that research? Like is that part of like the best way to present information to people to say this is how we went about this?
Millie (39:24)
Yeah, ⁓ if it's a research study ⁓ or a paper, they're going to do all of that. And even if it's a white paper, which is a paper that's not...
gone through a really specific process with the organization that has a bunch of peers who are not connected to review and make sure that everything's done in the correct way and that all things are thought through. If it's just a paper that's really well thought out of a research, they'll still also put in their methodology if they're of that kind of standard. I think...
⁓ Yeah, and I think they'll maybe simplify it in white paper situations, but I would say more often than not, most places, if they're trying to present stats of some sort, they're gonna give you all the details. There will be a link somewhere to a larger thing that you could really dig into. Whether or not that's always available to the public is another thing, because that was kind of my grip with the education space too, is that...
We spent a lot of time and money and effort and people power to answer really interesting questions and do things. But sometimes that just gets locked in a box that is behind a paywall of a paper of a source that people don't even have access to, that all you can read is the abstract, which is just giving you 10 sentences of a summary of the whole entire study, which is not enough. If people want to know more or other places can
Taina Brown she/hers (40:44)
Mm-hmm. A summary.
Millie (40:54)
like use that same study or methodologies to then expand upon. ⁓ So that's my whole other soapbox.
Becky Mollenkamp (she/they) (40:58)
So that's what I I
want to be mindful of time and I really wanted to make sure to ask this question about data as a product. And data maybe has always been a product. I don't know. But it seems to me anyway, in my very limited understanding that the age of online social media sites like Zuckerberg at all, realizing they were sitting on this goldmine of data, of our data, of user data.
Millie (41:09)
huh.
you
Becky Mollenkamp (she/they) (41:27)
all about who we are, our demographics, our psychographics, the things that we're interested in, what triggers us, all of that. They're sitting on all of this data that they've been collecting and that turning that into a monetizable product versus a subscription model or something like that. Maybe it's been there, but it doesn't feel like it's been there to the degree it has since they sort of realized, this is like a goldmine. We can sell to companies and to other people. And I'm curious, I don't know how much it relates to what you do, but
Millie (41:43)
Mm-hmm.
Becky Mollenkamp (she/they) (41:56)
What are your thoughts about data, like our data becoming a product? Has that changed? Does it change the way you do your work? Does it change the way people feel about data and about your ability to get information for people and that kind of thing?
Taina Brown she/hers (42:06)
you
Millie (42:12)
Yeah, I sigh deeply because as much as I don't want
Taina Brown she/hers (42:17)
you
Millie (42:18)
my data out there and being used against me, it already is. So does that mean that I should provide more if there's an option for me to not? No. So for example, that's why I forget what the acronym stands for.
Taina Brown she/hers (42:22)
It already is, yeah.
Millie (42:35)
but the GDPR, like that being a regulation in ⁓ Europe so that you can't sell, like you have to ask people on the site whether or not they will allow the cookies. And like that's what we see now. And that's because of the GDPR. It doesn't fully really apply to us really, but they have to put it on the sites.
But like it's nice that that is like even remotely an option. Whereas before, you know, all of this information is being pulled from whenever you access the internet.
And I don't think that's going to change. mean, cookies aren't as reliable because of that policy, but there are going to be like, always going to be other ways in which information is going to be pulled and used as like a product and like sold between companies. I think that's like always going to be the case. So it's just more so just like recognizing that as like a truth. And you know, you can use specific browsers like Brave.
or Firefox still does a good job of giving you extensions to help with tracking and things like that. But I unfortunately am in the camp of a pessimist of this nature, so I might not be the best person of just like, what can we do, or what is there? I've kind of just succumbed to like, you know, this is the reality of it.
Taina Brown she/hers (43:49)
Yeah, yeah, the scary thing. No.
Becky Mollenkamp (she/they) (43:50)
I just wonder also, sorry, I was just curious,
because does productizing data, does that affect data sets in some of the bias we're talking about? Or is it not really matter? It's just that now people are profiting off of the stuff that once you were able to access and you didn't have to, they weren't profiting off of it.
Millie (44:08)
Yeah, I think for me, at least my problem with it is more like the profitization in that I don't get any of that. Like you get to have this information. I don't even get a cut or a say. But I at the same time, you know, like I do, I do like the fact that there are things like I'm just going to come across things that are more relevant to me because they know of this like activity and information about me of like looking at that one site. And now I'm getting like the coupon to like potentially buy like, you know, I'll take a
Becky Mollenkamp (she/they) (44:33)
Yeah.
Millie (44:36)
I'll take a deal, like that's cool. So I can enjoy that aspect. But yeah, it's more so just like.
Becky Mollenkamp (she/they) (44:41)
Does it make your work easier?
Is there more data available to you now than there used to be? Because of all this data collection online? as people who are doing studies and wanting to find data for all the different things you're doing, is it easier to get more detailed data than you used to be able to? Or maybe it's not relevant.
Millie (44:45)
for ice products.
Mm-hmm.
It is.
I think, I mean, I don't personally like do like marketing data, but it does make it easier for like companies to sell like information or not information, but sell their product to like a population that would really buy it, which is great. And like even for like small businesses, I think that that's like, that's something that's like really helpful. And that like levels out the playing field. ⁓ They may not have that as much access to like the bigger like companies do, but I think that is like a good like benefit for for small businesses to like all.
have because like though there are businesses that do like help with pulling in that data and like giving that to companies so that they can like use it. I think yeah I think it's it's just something that will like continue to evolve and especially so with like the use of AI continuing to like scrape data but I don't know if it like I don't know if it like necessarily like makes like research or
⁓ things like that like easier yeah
Taina Brown she/hers (45:58)
What do you want people to still be thinking about after they watch or listen to this conversation? What's the takeaway for you here for people who may not necessarily interact with data on a day-to-day basis, from a research perspective, right? From a, is what I do for my job.
Millie (46:14)
Mm.
Yeah, I would say to be more curious about things. There is access information everywhere. And it doesn't have to be all doom and gloom either. One of the things that I think people, even if you're not trying to be a whole data scientist or data analyst,
most states have access to like have like open data sites. So like New York City has like NYC open data. And you just like this is information that your taxes are already paid for. So like you might as well use the resource. But you can like find things like one of the favorite data sets like the number of gray squirrels that there are like in locations and you can get that and NYC open data and that's open for everybody. It doesn't have to be like just like New York City.
But there's like information all around and like to be curious about it. Also to like be mindful of what you're like taking in and where those are coming from. But yeah, I love data. I love information. I think people should all be excited about it. I understand that like numbers is not everyone's thing, but it numbers are about you and like each data point anywhere is about a person and
It's just kind of like fascinating to me to like understand society through that like mechanism.
Becky Mollenkamp (she/they) (47:37)
And I'm the person who just goes and buys the first umbrella they see.
Taina Brown she/hers (47:40)
No
research. ⁓
Becky Mollenkamp (she/they) (47:43)
But I love the people who are going to analyze their umbrella purchase like good
for you more power to you. I think that's great.
Millie (47:46)
Yeah,
that's what the York wire, what is it? The site that they have, just a wire cutter, where they have done all this research for you. You can buy that first one. I will take that and then look for other ones to see.
Taina Brown she/hers (48:00)
You
Becky Mollenkamp (she/they) (48:01)
I'll be part of the data
that says how many were returned and then you would end up benefiting from all of that. Thank you so much for doing this, Millie. It was really great to meet you and have this conversation.
Millie (48:09)
Yeah,
thanks for having me. This was fun.
Taina Brown she/hers (48:10)
Yeah.
Becky Mollenkamp (48:12)
Thanks for listening to Messy Liberation. Will you please take a moment to subscribe, rate, and review wherever you listen? Taking that moment to do that will really help us reach a wider audience. Thank you so much for your support and for listening today. Until next time.