Really wonderful presentation, Allison! 

1. My experience is really in math, but I think the lessons hold true: learning that is authentic and gives ownership to the child, can be empowering, encouraging, and enriching. That's one thing that is particularly appealing about CS/programming, the ownership and authenticity is *almost* a given. We have much work to do to re-teach math educators (and redesign math textbooks) to make it as common in mathematics. 
2. The first thing that comes to mind is a recent site I found called LetToysBeToys.org.uk whose mission follows: "Let Toys Be Toys campaign is asking the toy and publishing industries to stop limiting children’s interests by promoting some toys and books as only suitable for girls, and others only for boys." I truly believe that advocacy work needs to happen from retailers, to toy makers, to advertisers, to families, preschools/daycares, and our entire community. Is that too much to ask? :)
3. I've often wondered what are the particular aspects of teaching and learning, not exclusively in STEM subjects, that cause a student to identify less with a discipline. In other words, what adult actions inhibit students' views of themselves as a mathematician, scientist, historian, author, storyteller, etc.? I think you've identified one key adult action here: gender-determined preferential access.

Keep up the great work! I have a little girl on the way and have spent a lot of time talking with our family about gendered toys and such. Probably too much, actually! Like all parents, we want her to have access to the full breadth of her possible futures and I don't want my toy-buying decisions to be a negative factor for her in any way. 

Hi Levi,

Thanks for your comments!

I appreciate the Let Toys Be Toys resource as well. When I gave talks about this research project, I showed what happens when you search on Amazon for "robot" within the categories of "Boys' Toys" and "Girls' Toys." There are more than 10 times as many results for Boys' Toys compared to Girls' Toys. But there has been progress from some companies (like Target) that no longer categorize toys by gender, and there are some great recent gender-neutral robot toys for children. We bought the Fisher Price Code-a-Pillar for our own preschool-aged daughter.

The development of identification is a fascinating question! We're about to start a large longitudinal project with elementary and middle school students in computer science classes, and we're very interested to see how their identification with computer science changes over time. We think sense of belonging may play an important role, especially for girls in STEM. It's hard to decide that me = STEM when you don't feel that people like you belong in STEM.

-Allison

Yes! We agree that it's so important for girls to believe they can do this and to identify themselves with science. We hope that starting early means that these experiences can add up over time, so that children build strong positive feelings about STEM.

Hi Jeanne,

Our sample was children in the Seattle area whose parents had signed them up to participate in research studies. So, this might be a group of children who had more familiarity with technology than average. We think it's very interesting that we still saw the typical gender difference in our control groups (with boys reporting more interest and confidence than girls), but that gap was much smaller in our robot group. 

We built into our research protocol a system of helping children who were having trouble programming, so that all children were ultimately able to succeed. We did keep track of how much help each child needed on each path. Overall, children programmed the path without assistance on 58% of paths, needed a little help with 21% of paths, and needed a lot of help with 20% of paths. There were no gender differences in how much help children needed. We also scored the number of “cycles to success,” the number of programs that children created and tested before successfully completing the path. Children succeeded on their first or second program for 64% of the paths. The children who needed more help programming did report lower confidence with robots, but enjoyed the robots just as much as the children who needed less help.

Thanks for your feedback!

Allison

Very interesting, Allison. As I was reading, I had a question that you answered at the end- about relationships between children who struggled more and their outcomes. I often hear people talk about "productive struggle" in learning and am very curious about it. I don't study it myself so I don't know much about exactly what that means or how people study it, but that's why I was wondering - does trying harder yield more confidence upon success or less? Of course, circumstances are always different. 

I'd be curious to know what happens in a duplicate study where you have a less selective group of kids. I don't doubt the gender gap will remain, but just curious about differences in confidence gains. We have done a study in Broward County Public Schools, looking at relationships between childrens' participation in integrated units that include CS and although we didn't see differences between treatment and control, we did see persistent gaps. 

Thanks for these interesting questions! "Productive struggle" sounds like a complex issue. Carol Dweck was my PhD advisor, so my first thought is that how children interpret struggle may depend on their mindsets. Children who have a growth mindset (and believe ability can improve) may be more likely to interpret struggle productively and become more confident. Children who have a fixed mindset (and believe ability is fixed) may become more frustrated by struggle and challenges, and become less confident. We're planning to incorporate mindsets into our future research on computer science education, so we may be able to give a more empirical answer in a few years!

Your second point is such a fundamental concern when we talk about any kind of motivation/achievement gap. Who benefits from educational interventions, and how much do they benefit? Is there a Matthew effect, where the already-successful become even more successful than others? How can we maximize benefits for the students who need the intervention the most without singling them out or stigmatizing them? In some of our other research on stereotypes about computer science, we've found that broadening the image of who belongs in computer science can boost girls' interest without decreasing boys' interest, which is encouraging.

Hi Linda,

Great suggestions! Yes, I agree that parents often don't even think about creating these opportunities for their children. It's also important to counteract the gender-STEM stereotypes, which can push girls away from STEM.

We've tried to reach out to families to increase awareness of this issue; here's an op-ed we wrote for the LA Times: "The gender gap in tech isn't set in stone"

Our research institute also has a wonderful Outreach division that works to share our research with families, educators, and policy makers. They've created a nice series of online modules, one of which is about early STEM learning.

We also think that integrating these opportunities into K-12 curricula for all children would be helpful. We're currently partnering with schools in Rhode Island (which has an initiative called CS4RI to put computer science in all K-12 schools) to see how computer science experiences in school affect students' interest over time.

-Allison

Hi Anna,

This is a tough and important question! Sapna Cheryan's lab has been looking at how social pressures can push girls away from STEM. She's found that college-aged women express more interest in computer science when you ask them privately than when you ask them publicly, but that's not the case for men. She had two ideas for how to deal with this issue; I'm not sure whether she has the results yet.

Her first idea was to give students private opportunities to explore their interests (away from peers). That might work for the parent you met (for example, the parents could send the daughter to STEM summer camp, away from that peer group).

Sapna's second idea was to work to change the stereotypes that create the pressure in the first place. Changing cultural stereotypes isn't something that can happen easily or overnight, but finding a friend or role model who does value STEM for their daughter could help those parents. 

-Allison

Thanks for your response! Those are great ideas. Another thought I had that I think lies somewhere in-between those two ideas is to engage the whole peer group in a stereotype intervention. So invite the child's group to your home for a movie night and ask them to pick a movie with a strong female scientist protagonist. Or to sit down and talk to them about their favorite female scientists and if they can't come up with any, ask them why they think that might be the case. My guess is that interventions of that kind will probably vary a lot in effectiveness depending on the peer group values/beliefs/organization, but it would be interesting to test whether in-person interventions of that kind have any effect on pre-teens/teenagers at the group-level.

Yes! And I think your suggestion includes an element that's critical for that kind of intervention--change has to come from the peer group themselves (supporting their autonomy and sense of fairness), not from an adult telling them what to think about STEM. There's a great study that convinced adolescents to make healthier eating choices by framing healthy eating as taking a stand against the food industry's efforts to manipulate children. Your conversation idea is similar--if they understand how society has created unfair structural barriers to women's participation in STEM, they might become more supportive of women in STEM.

Hi Merle,

We agree completely! Carol Dweck was my PhD advisor, and one of the next steps for our research is to see how teaching growth mindsets can help reduce the negative effects of stereotypes about STEM, and promote girls' sense of belonging.

Thanks for your comments!

-Allison

Hi Ashley,

Thanks for your comments! Yes, the Outreach division at our research institute has done STEM workshops with parents in the past, and they are currently working to get a grant that would allow us to do this in a much more systematic way. 

Our institute also has free online modules about child development that include tips and activities for parents, including one on Early STEM Learning.

-Allison

Hi Alexander,

We think that it is important to build CS into the elementary school curriculum, so that it becomes something that all children do, not just something "for boys." We're currently partnering with schools in Rhode Island, which has an initiative to put CS curricula into all K-12 schools. At the elementary school level, they're using the CS Fundamentals curricula from Code.org, Creative Computing from Scratch Jr./Scratch, and Launch from Project Lead the Way. We're excited to see how this kind of experience from an early age changes students' attitudes and stereotypes about computer science. In a pilot study last year, we didn't see many gender differences in interest in computer science, which suggests that these experiences may be an effective way to reduce gender gaps. 

We're planning to work most closely with the elementary school teachers in the next few years. It will be very interesting to learn how they're using the curricula and integrating it in their classrooms.

-Allison

Thanks for another great question, Anna!

I think preschool is the perfect age to start. There's a lot of research that preschoolers/kindergarteners can be successful with concrete, physical programming tasks. We bought our daughter the Fisher Price Code-a-Pillar when she was 3, and she definitely understood how to program it (although she wasn't quite ready for some of the more advanced ways to use it, like making it reach a target). These toys are great for building general cognitive skills like understanding patterns and solving problems. A good discussion of this is in Seymour Papert's book Mindstorms, about how computers can transform children's learning as they construct knowledge for themselves.

-Allison

I just saw this article this morning from NAEYC. It does a great job of articulating and extending the points I wanted to make, and has more great ideas for helping preschoolers build a foundation for learning how to code: Creating Play-Based Experiences

-Allison

Hi Pablo,

Thanks for your comments!

(I should mention first that our sample size for the robot group was small, and most children were from White and middle- to upper-middle-class backgrounds.) We found no gender or racial/ethnic differences in how much help children needed in programming. There was also no correlation between help needed and family income/mother's education level.

One of the things I really appreciated about the robots that we used (made by Adriana Moscatelli and Play Works Studio) was how the drag-and-drop programming was very intuitive for children. I think that's reflected in our findings about how confident children in our robot group felt after programming the robots. In our control groups, only about 70% of the children thought they were good at using robots, but almost all of the children in our robot group told us they were good at using robots.

-Allison

That's a great idea! Thank you!

Hi Rebecca,

Thanks so much for your comment. Our project was a great way to spark their interest, but there’s still so much to learn about how to hold that interest and keep girls excited about coding!

-Allison

Hi Mari,

Thanks for your comments!

The main goal of our research program is to find ways to boost STEM motivation in students, from preschool through high school. The gender gap in interest in computer science is such a large problem that many of our projects have focused specifically on this issue.

Research with young children is very fun—although it has its own challenges, like when a magic wand turns your lab table into a “frog”...

I’d love to talk more with you about your work as well!

-Allison