Whirley Pop DOE

Brooks Henderson
June 08, 2010
Written by
Ah, movie night! My wife and I agree that movie night, which we have in our house every Sunday, is our favorite night of the week.
Now, it’s not necessarily the movie that we enjoy so much. If we happen to watch a dud of a movie, it’s still a good night. I guess it’s just the time spent together that we enjoy and of course, the popcorn! We cannot have a movie night without those tender white morsels drizzled in butter and salted to perfection. A movie just wouldn’t be the same without it.

To enhance our favorite night of the week, we decided to experiment with our popcorn popping technique to try to maximize the delight of our taste buds and minimize the wasted unpopped kernels at the bottom of our bowl. There’s nothing worse than getting shorted on the pure enjoyment of eating popcorn.

Being experimenters, my wife and I have tried many popping methods, including the easiest√¢‚Ǩ”microwave popcorn√¢‚Ǩ”but we’ve found that nothing beats the taste and enjoyment we get from cooking it on the stove with the Whirley-Pop√¢‚Äû¬¢ popcorn popper. It’s pictured above.

The Setup: Factor and Design Selection

There are numerous variables to play with when using this stovetop popper. One question we wanted to answer was, “Do we really need to stand at the stove and stir the whole time it’s popping?” A second question arose from my parents’ recommendation that we pre-heat the pot by placing one kernel in the pan until it pops, and then adding in the rest of the kernels. We had never done it this way, so we wondered if it would make a difference. Lastly, we wanted to address whether spending the extra money for a premium popping corn was worth the price. So, we had three factors to investigate: Stirring, Pre-heat time, and Popcorn type.

We chose to do a full 23 factorial, testing all possible combinations. We also added three center points. The full factorial with center points is a good choice because it allows us to test for curvature. If curvature is detected, we can augment to a central composite design (CCD) to model the quadratic terms by adding a few extra runs to the original data. In all, we tested three factors at 2 levels:

A. Stirring, in fraction of time stirred from 0.0 to 1.0

B. Pre-Heat Time, in seconds from 0 to 360

C. Popcorn Type (categoric), from Cheap to Costly.

Because Factor C is categoric, the three center points we specified became six, three for each level of this categoric factor.

To facilitate using center points, we wanted to make as many factors numeric as possible. Factor A, stirring, was a little tricky. We didn’t want to just test stirring vs. not stirring, because that wouldn’t allow us to make a center point and test an intermediate amount of stirring. To make it numeric, we defined it as the fraction of time stirred. The low and high levels are self-evident, 0.0 is not stirring and 1.0 is constant stirring. At the center points, which required 0.5 fraction of time stirred, we stirred for 15 seconds, stopped stirring for 15 seconds, stirred 15 seconds...and so on.

The Results

We measured four responses for this experiment. Taste was rated on a 1 to 5 scale with 5 being the best. Texture was also rated on a 1 to 5 scale with 1 being hard and 5 being soft. We prefer tender popcorn, so a value of 5 was desirable here. We counted the number of unpopped kernels (UPKs) after popping and this was our third response. The final response was an estimate of the volume of popcorn in the pot after popping. Unfortunately, we didn’t have a graduated cylinder for measuring volume. That’s probably the only piece of superfluous kitchenware that we didn’t receive as a wedding gift. So, we just estimated the volume in the pot after popping. Given the wide range of our responses, it seems we had enough discrimination to draw some conclusions here.

Stirring and Pre-Heat both had a significant effect on taste and texture. In each case, it was best to go with constant stirring and no pre-heating; so much for our dreams of not having to stir at all. I guess Whirley-Pop knew what they were doing when they added that stirring mechanism to the pot. On the other hand, not having to pre-heat is a win for us. There is one caveat, though. Take a look at the stirring/pre-heat interaction graph for taste (Figure 1). At a stirring rate of 1.0 (right side of graph), it appears that no pre-heat (B-) will give the best taste. However, the error bars overlap slightly, so it can’t be deemed statistically significant. We do have enough information to go without pre-heating in the future, though. Even if we had just equivalent taste with and without the pre-heating, we can save time by cutting out this step. When things are statistically equivalent, go with the easier or cheaper option.

Another thing to note is the significant lack of fit detected by our center points. Look at the green dots in Figure 1. These are the results at our center point treatments using the expensive popcorn. The true data points lie above the lines predicted by our model. It appears we may be able to get better tasting popcorn if we only stir half of the time or some intermediate proportion. I like the sound of that: less work and better taste! The ANOVA summary table also showed that there is significant curvature with a p-value of 0.0058 for the adjusted model.

The significant curvature suggests that we can better fit the data by adding a quadratic term, either A2 or B2, but because we only did a factorial design, we don’t have enough data to determine which term is needed. We will have to augment the current design with more points to get a CCD for response surface methods (RSM), which will tell us which term is correct.

For UPKs, the only factor found to be significant in the design space was factor A, stirring. More stirring reduced the number of unpopped kernels significantly (from a mean of 73.9 UPKs to only 2.8 UPKs). The volume of popcorn in the pot after popping varied mostly due to the stirring and pre-heat factors. Just like with taste, more stirring and no pre-heating was the best combination, giving us the highest volume of popcorn.

What did we learn?

Constant stirring (a level of 1.0) with no pre-heating gave the most desirable responses for taste, texture, and volume of popcorn. The brand of popcorn didn’t make a statistically significant difference over the range tested, but it did appear to have some effect. We should include it in further studies. The analysis also detected significant curvature for the taste response that would suggest that we can get a better taste rating if we stir for an intermediate proportion of time. However, constant stirring reduced the number of UPKs. To decide whether we should use constant stirring to reduce UPKs or an intermediate level to possibly increase taste and texture, we’ll have to upgrade to a response surface design so that we can model that intermediate proportion of stirring. If we want to keep the same factors and levels, we can simply augment the current design with a few more runs to get a CCD. However, we may want to throw the other known factors (time and temperature) that we didn’t test into the final characterization and optimization study. Stay tuned for Part II.

√¢‚Ǩ”Brooks Henderson, DOE Consultant, Stat-Ease, Inc. (

This e-mail address is being protected from spambots. You need JavaScript enabled to view it

Brooks is the newest trainer/consultant at Stat-Ease, Inc. Take a look at his biography at


0 #2 Brooks Henderson 2010-09-02 10:46

Thanks for your your comments on my article. It was a fun experiment and went well with our normal popcorn eating ritual anyway. I've never tried peanut oil. Does it taste a little nutty?

There are a lot of other things we could do, but we're pretty happy with our yield and taste now.

I've never tried the popcorn straight from wabashvalley, either, but I do like the Orville Redenbacker. I think the fact that they have a resealable and thick plastic container helps, since I've heard from others that moisture content in the kernel's has a big effect on the amount of unpopped kernel.
0 #1 Benoit Bouchard 2010-06-17 06:30
OMG! I find this test so funny!
I've used a Whirley Pop for about a year and also found that the oil used for cooking (+qty) makes a huge difference taste-wise, so you have do redo the whole thing again!!! Haha!

Ok, now back to serious stuff (...) : I prefer pre-heat+1tbsp of peanut oil. The only quality kernel I can find in my region is Orvil Redenbacker (in the funny shaped container). I have yet to try the ones offered on Wabash Valley Farm's website (

Thanks for putting a smile on my face.

Add comment

Security code

Subscription Centre

New Subscription
Already a Subscriber
Customer Service
View Digital Magazine Renew

Latest Products

Popular Articles

Latest Events

No events