dbrawner
Private, 1st Class
- Messages
- 49
I'm having a challenge trying to interpret the broker evaluation data.
Since there is no way to sort the data on ratings AND reviews, one can not get a clear picture of which broker is really on top.
If we sort by rating, a dozen brokers are rated 5 star BUT... the these have too few reviews to be statistically significant by reasonable standards. In other words, having only 7 reviews rating a broker at 5 stars does not give me a warm and fuzzy feeling.
On the other hand, it seems the most reviewed sites (432 to mid 100's reviews) are mostly sub to average at 2-3 stars.
So, statistically, a high rating from a few reviewers is probably simple bias, while a statistically significant number of reviews plummets the ratings to, well, just average or slightly below. (actually this could be expected but it doesn't help a broker selection process).
Am I missing something or does the industry have a real problem with retail brokerages being perceived as sub-par?
Is there a statistician out there who can make sense of the broker review data?
Frankly, I have two issues with the broker reviews;
First, the reviews themselves and the ratings given. I'll wager that fewer than half are meaningful from the standpoint of the rating system (https://www.forexpeacearmy.com/fore...riptions-forex-peace-army-review-ratings.html) or the qualifications of the reviewers. Certainly, with the exception of some very specific examples, most reviews were too generic in content to support the rating provided. Also it is difficult to determine who, or how many, reviewers weighed in more than once, or who may have had a personal agenda (i.e. bias) for rating a broker one way or another.
My Second Challenge; the evaluation and presentation of the broker ratings. The top reviewed broker, with 432 reviews, received a 2.5 star rating, it's next competitor, with 341 reviews was rated 3 stars... on down to a rating of 3 stars from 215 reviews. In fact, of the top 6 (those having more than 200 reviews) the average rating was only 2.75 (the median being 3.0).
Further analysis revealed;
12 brokers with 5-star ratings received an median of only 4 reviews (high 7, Low 3).
17 brokers with 4.5-star ratings received an median of 5 reviews (high 34, low 3).
18 brokers with 4-star ratings received an median of 5 reviews (high 137, low 3).
35 brokers with 3.5-star ratings received an median of 13 reviews (high 197, low 3).
18 brokers with 3-star ratings received an median of 8 reviews (high 341, low 3).
and 368 brokers received a rating less than 3-stars with the best of the lot receiving 2.5 stars on 341 reviews.
Based on these data, statistically speaking I wouldn't risk my money with any of the 607 brokers reviewed.
Those having a significant review count are sub to average at best, and those with commendable ratings have an insufficient review count for the rating to be considered reliable.
Am I missing something???
Since there is no way to sort the data on ratings AND reviews, one can not get a clear picture of which broker is really on top.
If we sort by rating, a dozen brokers are rated 5 star BUT... the these have too few reviews to be statistically significant by reasonable standards. In other words, having only 7 reviews rating a broker at 5 stars does not give me a warm and fuzzy feeling.
On the other hand, it seems the most reviewed sites (432 to mid 100's reviews) are mostly sub to average at 2-3 stars.
So, statistically, a high rating from a few reviewers is probably simple bias, while a statistically significant number of reviews plummets the ratings to, well, just average or slightly below. (actually this could be expected but it doesn't help a broker selection process).
Am I missing something or does the industry have a real problem with retail brokerages being perceived as sub-par?
Is there a statistician out there who can make sense of the broker review data?
Frankly, I have two issues with the broker reviews;
First, the reviews themselves and the ratings given. I'll wager that fewer than half are meaningful from the standpoint of the rating system (https://www.forexpeacearmy.com/fore...riptions-forex-peace-army-review-ratings.html) or the qualifications of the reviewers. Certainly, with the exception of some very specific examples, most reviews were too generic in content to support the rating provided. Also it is difficult to determine who, or how many, reviewers weighed in more than once, or who may have had a personal agenda (i.e. bias) for rating a broker one way or another.
My Second Challenge; the evaluation and presentation of the broker ratings. The top reviewed broker, with 432 reviews, received a 2.5 star rating, it's next competitor, with 341 reviews was rated 3 stars... on down to a rating of 3 stars from 215 reviews. In fact, of the top 6 (those having more than 200 reviews) the average rating was only 2.75 (the median being 3.0).
Further analysis revealed;
12 brokers with 5-star ratings received an median of only 4 reviews (high 7, Low 3).
17 brokers with 4.5-star ratings received an median of 5 reviews (high 34, low 3).
18 brokers with 4-star ratings received an median of 5 reviews (high 137, low 3).
35 brokers with 3.5-star ratings received an median of 13 reviews (high 197, low 3).
18 brokers with 3-star ratings received an median of 8 reviews (high 341, low 3).
and 368 brokers received a rating less than 3-stars with the best of the lot receiving 2.5 stars on 341 reviews.
Based on these data, statistically speaking I wouldn't risk my money with any of the 607 brokers reviewed.
Those having a significant review count are sub to average at best, and those with commendable ratings have an insufficient review count for the rating to be considered reliable.
Am I missing something???