It appears not. We measured the “Growth Mindset” of 218 students using the following questions. In this blog I do a more detailed breakdown of results.

I would like to thank Panorama Student Surveys for publishing their measures of Social-Emotional Learning survey and allowing schools to use the questions.

The correlation between the “Academic Achievement” and other student data sets are shown.

As you can see the correlation between growth mindset and academic achievement at our school is quite low.

Academic Achievement | |

Work habits (Progress Score) | 0.84 |

Task submission | 0.70 |

ARCC Awards | 0.46 |

Attendance | 0.33 |

Co Curricula Points | 0.29 |

Growth Mindset | 0.10 |

Out of Uniforms | -0.28 |

Lates | -0.39 |

Misconducts | -0.40 |

For those unfamiliar with what correlation means here is a quick summary.

Correlation coefficients are in the range of +1 to -1.

If two sets of data have a correlation of 1 it means there is a direct correlation between them. When one data set gets bigger so does the other. An example of this would be the relationship between the mass of water in a bottle vs the volume of water in a bottle. When one goes up, so does the other in a direct relationship.

If a correlation coefficient is -1 it indicates as one set of data gets bigger the other will get smaller. An example of this would be the relationship between the number of cars on a freeway and the average speed of the cars. When the number of cars goes up the speed goes down.

If the correlation coefficient is 0 it means the two sets of data appear to have no effect on each other at all and may be unrelated.

The highest correlation we see in our school is between Academic Achievement and Work Habits. Progress score is our measure of work habits. At our school progress score is calculated every 5 weeks and is made up of teacher assessments of Effort, Behavior, Organization, Task Submission and Academic Progress. It is interesting that Effort and Behavior are specifically mentioned in the Growth Mindset questions but the students who believe they can significantly change these don’t perform better than those who feel these are not possible to change.

Task Submission has a high correlation with achievement. Task submission is a percentage measure of the rate of task completion. It is clear that “Those who do the work, do the learning.”

ARCC awards are given by teachers when students display the College Values of Achievement, Respect, Commitment and Community. There is a positive correlation between ARCC awards and achievement.

The correlation between Attendance and Achievement is not high. But the relationship between achievement and attendance is not straightforward. For students with attendance above 95% there is almost no correlation between Attendance and Achievement. Students with 95% attendance seem to do as well as students with 98% attendance who do as well as students with 100% attendance. Once attendance drops below 90% achievement begins to decline significantly.

Co-curricula points measure student’s participation in co-curricular activities like sport, drama, music, debating and leadership positions. There is a small positive correlation between participation in co-curricula activities and Academic Achievement.

Growth mindset comes next on the list. A correlation of positive 0.1 is almost no correlation at all. Students with a growth mindset seem to do just as well academically as those with a fixed mindset.

There is a negative correlation between achievement and being Late to Class, Out of Uniform and Misconducts.

I would love a dollar for every time I have had the discussion with students who are frequently out of uniform and they have made the point, “Being out of uniform doesn’t affect my learning”. Of course on one level they are correct. Correlation is not causation. Being out of uniform does not prevent a student from learning but I can now point out that “On average, students who are frequently out of uniform don’t do as well at school as those who are in uniform. There is a negative correlation between being out of uniform and Academic Achievement.” Not sure this will change their hearts and minds but it will be a teachable moment for me. I will launch into a fascinating explanation of correlation and causation and the nature of statistical analysis. They won’t want to be out of uniform again after I walk them through that one a couple of times.

Students who are frequently late to class, frequently out of uniform and have many misconducts don’t do as well at school. That is hardly a surprising finding.

So what do we do about all that?

Is Dweck’s Growth Mindset the next “learning styles” educational myth that is not supported by empirical evidence?

We all remember studying “Learning styles” at teachers college. Most of us who have been teaching for longer than 10 years can still remember the “talking head guru” Professional Development Day presenter who talked about activities that help the kinesthetic learners and the auditory learners and visual learners. We raced out and gave our students VKA Learning Styles Surveys to identify their preferred learning style. The problem was that delivering material to them in their preferred learning style didn’t actually improve learning.

The consensus view on learning styles now appears to be:

Do students have a preferred learning style? Yes

Does delivering content to students in their preferred learning style improve learning? No

I am not suggesting we put “Growth Mindset” in the same category as Learning Styles.

Much of Dweck’s evidence for Growth Mindset was done in a “pre and post test” methodology. She looked a group of students with similar mathematical ability and tracked their progress through Maths in Year 7 and 8. She found those with a “growth mindset” improved more than those with a fixed mindset, especially when confronted with difficult problems. The other studies that used some forms of control groups and experiment groups had the control group doing a normal program. The experimental group did the normal program plus some work on developing a growth mindset. The group that did the work on growth mindset not only developed more of growth mindset but they also performed better academically as well relative to the control group.

Should we be surprised that many students with a “fixed mindset” can do well at school? We have all taught students who when asked why they do well on a task say, “Because I’m smart.”

I am curious.

I want more information.

Is there a correlation between Academic Achievement and Growth Mindset? We certainly didn’t find one here. Was my method flawed? Are these mindset questions not a good measure?

My challenge to other schools: Measure student’s growth or fixed mindset and see if it correlates with academic achievement.

Tell me what you find.

[…] Results are here. […]

LikeLike

As much as you praise statistical analysis here and point towards conclusions, your own statistical analysis here is pretty weak. Have you ever thought of the idea that academic achievement itself might influence most of your variables? Might there be other confounding factors? Raw correlations without controlling for anything is for the trashcan. You cannot rush to conclusions about the growth mindset here.

(cum hoc ergo propter hoc)

LikeLike

Hey Rob,Just watched the videos. Very interesting – and inspirational – as ever! Thanks for all the time and efforts you put in to your vids and your posts as well!Another great toolbar add-on is Web Developer – very handy for checking pages that you’re building, and also for &#ur&0;deconst28cting2#8221; other people’s web pages that look interesting!Anyway, best wishes, PhilGD Star Ratingloading…

LikeLike