Subscribe Bookmark
anne_milley

Staff

Joined:

May 28, 2014

Baby not sleeping well? Design of experiments to the rescue!

Inna Ben-Anat of Teva Pharmaceuticals designed experiments to figure out what was affecting her daughter's sleep.

Inna Ben-Anat of Teva Pharmaceuticals used experiment design to help figure out what was affecting her daughter's sleep.

DOE (design of experiments) certainly helps make the world a better place in big important ways, but it can also add to your personal quality of life. In addition to the practical and clever experiment designs that developers Ryan Lekivitz and Melinda Theilbar have recently discussed here in the JMP Blog, here is another one we hope will inspire you.

This DOE example comes to us from an enterprising mom who also happens to be Director of Global Quality by Design Strategy and Product Robustness at Teva Pharmaceuticals, Inna Ben-Anat. She has used this in one of her DOE training sessions and kindly agreed to let us share her findings.

Her 6-month-old daughter was waking up as many as eight times at night, and Inna wanted (and needed) to get better sleep. Her main response was how many times she woke up at night using these factors over a 19-night study:

  • Length of nap(s) during the day
  • Last feeding: fruit/veg
  • Bath: yes/no
  • Total ounces of night bottle feeding
  • Waking: wet/dry
  • Mattress: old/new
  • Time of bowel movement
  • Screen Shot 2015-11-17 at 2.25.17 PM

    She found that the most significant factor was the mattress! She had bought a new mattress with less rounded edges so her daughter wouldn’t slide to the sides. She also found that no bath, no leaky diaper and night feeding were important factors as well. Here's Inna’s humorous conclusion: Have the baby hungry and not clean. But seriously, the new mattress made the difference!

    She even has success to report for her “generation II” implementation for her now 8-month-old son’s feeding and sleeping schedule for his first weeks in daycare. These are Inna's control charts below:

    Screen Shot 2015-11-17 at 1.49.27 PM

    Screen Shot 2015-11-17 at 1.47.44 PM

    They called him a cat-napper in daycare, and we can see why. We can also see that the first day of daycare, he went on a hunger strike, but then his appetite took off on subsequent days. It looks like her son's transition to daycare was successful, but kudos to Inna for collecting some data "just in case" there were any issues.

    Thanks to Inna for sharing these great examples. I applaud her practical problem-solving skills!

    For even more and varied DOE examples, check out the more than 100 example experiments used in teaching at Curious Cat Management Improvement -- and please feel free to tell me about your own!

    Want to start designing your own experiments? Start with our on-demand webcast on basic design of experiments.

    6 Comments
    Community Member

    François wrote:

    Spendid DOE Inna!

    Is woke up at night time at 0 now?

    Community Member

    Inna wrote:

    Hi François!

    Letâ s just say significant improvement was achieved :-)

    Now itâ s more about how do I make them go to sleep from the first placeâ ¦.once they are out-all is good ...

    With JMP making sense of all this data around us is so easyâ ¦

    Hope all well with you.

    Community Member

    Sun wrote:

    Inna, so surprised you had the energy to do this experiment.

    Can you do another replicate block (next child)? :-)

    Community Member

    Sun wrote:

    I also shared on our QbD LinkedIn forum:

    https://www.linkedin.com/groups/848287/848287-6078569165155823618

    Community Member

    Inna wrote:

    Hi Sun, I would rather wait for findings validation from external source...any volunteers? :-)

    Thanks for sharing on QbD community, There we can say the project is now at scale-up stageâ ¦:-)

    Community Member

    Louis Valente wrote:

    Inna,

    You motivated me to go out and replace my mattress. If it's good for baby it's gotta be good for me:)

    Lou V