The descriptor formats and correlated drop in results refer to how they collect their data, not the analysis. When they perform their analysisusing binary, tertiary, distributive etc, they find that the method of analysis does not affect the results. Read the paper carefully and I think you'll eventually see what I mean. It's a tough going paper I know.
And I have read it many times over, and am all too familiar with it, thank you.
For example, on Table 5, the remote viewers record their impressions using the distributive method (10 options per question). PEAR then treat this data as if it were collected using either binary, ternary and various distributive methods. From the table, we can see that data collected and analysed using the distributive method got no results. However, if the significant results of the previous binary experiments (tables 1-3) were due to the analysis method then we would expect to see the same artifactual result when the distributive data is treated in the same way. But this does not happen. How do you account for this?
In simple terms, the reduction of distributive data to binary data causes "clumping" of results - rounding errors if you like, which leads to artifacts. Binary data, no matter how it is obtained, cannot then be redistributed to obtain finer granularity for analysis.
For example, you cannot reasonably extract a point value in the range 1 to 10 out of Yes/No base data. Such information can only be obtained at the source, which PEAR did not do in some of their studies. However to aggregate their data they needed to massage it into some common "format", and the lower granularity was it.
I think their use of the phrase "increased emphasis on objective quantification" is not meant to imply that their data collection methods gradually reduced artifactual subjective bias. They were just using more options on their descriptor sheets.
I don't believe it refers to data gathering at all.
If you read the paper, it refers specifically to the (rather pathetic) apologetics used by PEAR to justify claiming there ARE results even when their own data and analysis shows otherwise. They are saying that the results they "wanted" were only present when the data was viewed subjectively. That is, someone is of a baseless opinion that it is there. Which is the direct antithesis of the objective scientific method, don't you agree?
I thought one of the main points of the article was that the drop in results were due to progressive removal of subjective bias from their analysis. I do not think the results show that at all. Thats not to say the PEAR RV experiments contain no methodological flaws, cos they cleary do such as inadequate randomisation of targets. I just think it does the sceptical position no good by introducing errors like that in the article because it obscures the real issues. And yes, the language of the PEAR people is, ahem, individual, but I don't see that as a real problem.
There were a number of issues facing PEAR when they went to analyse their data:
First, the issue of poor data. As mentioned above and also commented on extensively elsewhere, the quality of base data being used ranged from poor to mediocre, with glaring irregularities being included. The data obtaining methods varied considerably over time (the various studies ranged over 25 years!), and produced disparate data sets. The study methodologies were not particularly similar either, nor were they particularly well conducted. It seemed that ANY study that had "remote viewing" as its premise was included (see other comments above).
Second, the issue of poor arithmetic - a mechanical issue, basically. It stemmed almost entirely from trying to lump the sets of disparate data together and trying to draw viable conclusions from them. As mentioned above, to do this they had to discard some "information" in the process (reducing variant data to binary, etc). This was also the subject of the extensive criticism from elsewhere, and the main reason why they went to the lengths of progressively reviewing their analytical techniques (i.e. "improved recipes", as they called it).
However credit where credit is due - the PEAR analysts have not fudged the analysis results, as far as I can determine anyway, and were honest in their reporting of same. That is, they were objective in as far as this went.
But it is the swing away from being objective to being subjective that is the major flaw.
