2
$\begingroup$

enter image description here

The box-plot above represents the number of programming tasks performed correctly in 30 minutes during an experiment by me. Each box represents an age group:

  • p1 = 6 - 7 years old
  • p2 = 7 - 8 years old

  • ...

  • p6 = 11 - 12 years old

  • s1 = 12 - 13 years old

  • s2 = 13 - 14 years old

  • ...

  • b1 = 16 - 17 years old
  • b2 = 17 - 18 years old

When I see the results, I am surprised by the difference between p4, p5 and p6. The set of tasks was almost identical by age. The only difference is that by increasing the age they have some less start tasks. I think it may be due to the ability of reading comprehension and motor development ...

Are there more factors that can influence?

UPDATE

enter image description here enter image description here

enter image description here

enter image description here

|improve this question
$\endgroup$
  • 1
    $\begingroup$ What do you mean by "less start tasks"? Fewer tasks initiated/attempted? Your box plot only shows completed tasks... $\endgroup$ – Fizz Aug 21 '18 at 2:17
  • 3
    $\begingroup$ Please note that this question has been double posted. $\endgroup$ – Ben I. Aug 21 '18 at 5:36
  • 1
    $\begingroup$ @Fizz fewer beginner-level tasks (assigned) $\endgroup$ – Marina Aug 21 '18 at 13:56
  • 1
    $\begingroup$ Ah, I think I might have realised, is this related to your previous question psychology.stackexchange.com/q/20437? $\endgroup$ – Chris Rogers Aug 22 '18 at 21:04
  • 2
    $\begingroup$ @ChrisRogers Yes, it's the same study $\endgroup$ – Marina Aug 22 '18 at 21:30
2
$\begingroup$

Are those reliable differences? Without knowing what confidence intervals you plotted it's difficult to estimate visually (95%? ±1 SD?), but I wouldn't be surprised that there are no statistically significant performance differences between those levels. Keep in mind that if this test emerged only after seeing the data, rather than a priori based on theory, then that changes how to interpret the significance of such tests and makes them less trustworthy:

HARKing: Hypothesizing After the Results are Known http://journals.sagepub.com/doi/abs/10.1207/s15327957pspr0203_4

The main finding seems to be an increase with age, possibly non-linear. Thresholding could be due to the task complexity.

|improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.