https://www.cnn.com/2020/11/26/europe/astrazeneca-oxford-vaccine-data-concerns-intl/index.html 

Cut and paste snippets from the article (yeah, I'm cherry picking)

AstraZeneca said on Monday that study participants in the UK had been given two different courses of the vaccine.

The drugmaker, which developed the vaccine with the University of Oxford, did not however, at the time, explain why they used two different dosing regimens or why the size of one group was significantly smaller than the other.

In one group, 2,741 participants received a half-dose of the vaccine and then a full dose at least a month later. This group was 90% protected against Covid-19.

In the second group, 8,895 participants received a full dose followed by another full dose at least a month later. This group was only 62% protected.

That's why AstraZeneca says their vaccine is 70% effective, on average.

On Tuesday, Mene Pangalos, AstraZeneca's Executive Vice President who heads up non-oncology research and development, first explained to Reuters that a lab error was the reason why some volunteers had received a smaller dose -- the dose that proved to be 90% effective. "The reason we had the half dose is serendipity," Pangalos said, adding that researchers had "underpredicted the dose of the vaccine by half."

Oxford elaborated further on Thursday, saying in a statement to CNN that a "difference in the manufacturing process" had led to the error.

Speaking to the Wall Street Journal on Wednesday, Menelas Pangalos AstraZeneca's Executive Vice-President said, "The mistake is actually irrelevant.

"Whichever way you cut the data -- even if you only believe the full-dose, full-dose data... We still have efficacy that meets the thresholds for approval with a vaccine that's over 60% effective," Pangalos said, according to the WSJ.

The US Food and Drug Administration requires a threshold of at least 50% efficacy. It is not clear if the FDA will authorize emergency use however.

In a call with reporters on Wednesday, the US vaccine czar Moncef Slaoui said they were reviewing AstraZeneca's vaccine data.

He said that group that got the mistakenly lower dose that yielded the 90% efficacy had been a younger group, with no one older than 55. That could potentially affect the strength of AstraZeneca's findings, given that young people typically produce stronger immune responses to vaccines.

My cliff notes:
Ooops, we kinda messed up on the study. Call it a lab error, serendipity, manufacturing error... whatever you want. Just minor details getting in the way. Let's not focus on silly details like testing methods. We're in a hurry here! Keep your eye on the prize. We just need numbers that show the efficacy still meets the approval requirements of 60%. Besides, the FDA says they will be happy with a 50/50 coin flip anyway.

Oh, the 90% protected group? Yeah... well... they were all from a younger batch of humans. We all know that age plays no role in this thing so we just lumped them in with the 60% protected group. It all averages out