First, the limit for foodstuffs of Cs-134 and Cs-137 is 1,000 Becquerels per kilogram according to the linked page.
The student (or the reporter) says that the IEAE limits are "1,450 counts in 10 minutes" (that seems a strange unit to me), which works out at 0.4027777 counts per second, or 0.4 becquerels. Thus, for her to claim that as the limit, she must be measuring 2.5 tonnes of seaweed!
However, a bit more searching turned up this document from Japan - on page 3 it says:
ICRP : alpha emitters 0.37 Bq/cm2
: beta and gamma emitters 3.7 Bq/cm2
JAERI : alpha emitters 0.4 Bq/ cm2
: beta and gamma emitters 4.0 Bq/ cm2
So, we have our 0.4 Bq figure, perhaps, but per square centimetre, but both Cs's are gamma emitters...
To conclude, I can only say that I suspect she has made some sort of false assumption or misreading of nuclear contamination data somewhere along the way.
[I'm also exceptionally suspicious of the seaweed being from China - who knows what other pollutants might be in that]
UPDATE: I believe this is a better link that includes a photo of her Geiger counter, which is a bog-standard one, not a specialised food-measuring one. She also says:
Radioactivity is measured in becquerels (Bq), and 0.5 Bq per square
centimetre is widely considered an actionable level of contamination.
Delacruz said one Bq is equivalent to 1,450 counts over a 10-minute
period, and many of her samples tested well over this amount.
There we go. One Bq is one count per second, and last time I checked there were 600, not 1,450 seconds in 10 minutes. Furthermore, she is using a surface contamination figure, as I suspected, so I wonder if the experimental procedure was to cut out a 1 cm square of seaweed and sit it on top of the detector for ten minutes?
My updated conclusion: yes, false assumptions leading to a meaningless experiment. Junk. Here's a detailed description of how to measure surface contamination properly.