A group of infants are identified as having low IQs, although it is admitted that infant IQ scores have low reliability. Five years later, most of these children have higher IQs than before. What is the most reasonable and most parsimonious conclusion?

A) The test that was given at the later time must have been easier than the first test.
B) The increase is due to random fluctuations in unreliable data.
C) IQ tests get better and better each year.
D) These children must have profited from some special experience they received during the 5-year delay.

The fact that infant IQ scores have low reliability suggests that the initial test results may not have been accurate reflections of the infants' intelligence. It is also known that IQ scores can increase with age and experience, so it is possible that the infants simply matured and acquired new skills in the five-year period. Option A is less likely since IQ test results are standardized to account for changes in difficulty over time. Option C is not supported by research evidence. Option D is possible but not necessary to explain the results. Therefore, the most reasonable and parsimonious conclusion is that the increase in IQ scores is due to random fluctuations in unreliable data.