It's possible that the information provided by the sources does not constitute a contradiction. In order for information to be considered contradictory, it must come from at least two independent sources that cannot both be correct.

o When thought through carefully, though, everything becomes crystal evident. Truth and reality are the focus of scientific inquiry. A scientific experiment is a query about the world, answered using the best available means of observation. Your empirical findings are a window into the world and reveal something significant about the subject at hand. But the raw data is also a byproduct of the process, and there are always certain to be limitations, biases, and deceptions in any methodology.

 Analysis performed on raw data has the potential to further muddy the waters, since it has a tendency to distort the image and toss away valuable information. The interpretation and what is seen are both susceptible to biases introduced by the observer, in this case the scientist.
 

Many important details are brought out in the preceding description. Experiment outcomes may be interpreted in a variety of ways. Remember that numbers aren't lying to you. When the raw findings pan out, they will reveal a lot. Scientists are tasked with providing explanations and explanations to their peers.

Okay, let's get to the question. Different people are certain to have different opinions on any given topic. When asked to describe the identical incident, any two persons will present unique perspectives. You can count on this. What we mean by "replication" is two independent but similar occurrences in the scientific community. Every storm is unique. If the findings are correct and truthful, then they cannot be contradictory. They should either work together or be separate (i.e., not the same thing). 

When the analysis is different, when interpretation alters them, the differences are accentuated. In order to completely grasp the findings of a study, one must first examine the procedures used to collect the data. Any little change in procedure might have a significant impact.

All too often, scientists are not forthcoming with their processes and data. The inability to think clearly is exacerbated by bad scientists. This adds the trickiest variable of all: the egotism of individual humans.

To sum up, the discrepancies that have been seen are normal and to be anticipated. Statistics can never be questioned because they never lie. The issue of the seeming contradiction must be addressed where it is most immediately experienced by the individual.

When a statement claims both of its opposing premises to be true, it is said to be contradictory. My only sibling is envious of me, for instance. Words like "contradict," "contrary," and "antithetical" all have roots with "contradict," "contradict," and "antithetical," respectively.

I understand that blogs are meant to be timely; otherwise, what's the sense of making posts that will eventually be forgotten? However, every once in a while I stumble onto a resource from a year or two ago that is so helpful that I feel compelled to pass it forward. This story from the Science Times section of the New York Times is a prime example. 
 

It's a shining example of how a journalist may perform outstanding work in demystifying scientific concepts for the general reader.
There was a really helpful issue of NYT Science Times released in 2008 titled "Decoding Your Health." This issue was a direct response to the plethora of health-related resources accessible to consumers today, including the internet, print media, and medical professionals themselves. If you're having trouble "decoding" all this information and sorting out what's valuable from what isn't, the articles will be a big help.
I was particularly interested in one piece titled "Searching for Clarity: A Primer on Medical Studies." I've seldom come across such a well-organized discussion of how medical data builds to inform recommendations for patients.
                                                                                                                                 

They choose a case that represents a common customer problem as an illustration. It was widely believed in the 1990s that the antioxidant beta carotene, contained in foods like carrots, squash, apricots, and green peppers, may be beneficial to health. Some animal and observational studies also seemed to indicate that beta carotene protected against cancer, lending credence to this theory. The supplement industry benefited greatly from the demand for beta carotene pills.
Then three big, well-executed clinical studies were published, in which participants were randomly allocated to receive either beta carotene or a placebo. These results indicated that beta carotene supplementation does not protect against illness and may increase the risk of cancer in certain individuals.
Frankie Avalon was a popular adolescent idol in the 1950s, with songs including "Cupid," "De-De-Dinah," and "Tuxedo Junction"; if you were a TV viewer at the time, you may recall seeing him in an advertisement. The article states that he was staring at a mountain of papers proclaiming "beta carotene works," and a mountain of papers representing the three research demonstrating its ineffectiveness. 

Takeaway: It's unclear who you can trust.
Clinical trials, that's the solution. The essay spells it out clearly, demonstrating how three basic concepts contribute to a more conclusive study:
As the saying goes, "the groups you are comparing must be the same except for one factor — the one you are studying."
 

The results of a study are more credible when they are based on a large sample size. They make a valid argument that scientific research don't provide a single figure but rather a range of results (such as "you have a 10-20% reduction in risk"). More extensive sampling increases confidence.
The conclusion should also make sense. The conclusion shouldn't be made out of thin air; there should be facts to back it up.