Specifically the study showed "statistically significant differences between churches of 100 or fewer adult attenders and churches of 1000 or more adult attenders." In fact, the only issue covered where no real difference existed was whether the person had prayed during the previous week. Here is some of the information from the article at Barna.org.
On all 9 of the belief statements tested, attenders of large churches were more likely than those engaged in a small or mid-sized congregation to give an orthodox biblical response - e.g., the Bible is totally accurate in all the principles it teaches, Satan is not merely symbolic but exists, Jesus led a sinless life, God is the all-knowing, all-powerful creator of the world who still rules the universe, etc.
On seven of the eight behavioral measures, attenders of large churches were substantially more likely than those of small churches to be active. (These included behaviors such as attending church in the past week, reading the Bible in the past week, volunteering at their church in the past week, etc.) The average difference related to these seven behaviors was 17 percentage points.
It was also shown that larger churches were more likely to have college graduates, wealthy attenders, and attenders/members with children under 18. Adults in these Protestant mega-churches were also more likely to vote Republican.
Another interesting point was that House Churches were not following the trend of other small congregations.
The religious beliefs and behaviors of people who attend house churches, which average about 20 adults in attendance, are more similar to the results for large conventional churches (i.e., more than 500 adults) than they are to the outcomes among those who attend small conventional churches (i.e., less than 50 adults).
So head over to the Barna Group, read the report and come back here to discuss.