I have some experience with the creation of test instruments for use in a Human Resource setting. As I engage more deeply with the TBI check list, I have developed concerns with its validity. For one thing, I doubt the items have been subject to factor analysis. There are a number of question bank items which are fairly similar in terms of the response they attempt to elicit.

The scoring instructions request each item be scored on the experience of the past 24 hours. This also strikes me as problematic. I presently live on my own. It is highly unlikely I will experience anyone giving me verbal instructions in the next 24 hours, or within the past week, or year. Even were I living with a partner, I believe it likely there would be extended periods with no form of verbal instruction. Some of my responses are derived from interactions arising in the context of shopping in a crowded and noisy venue. Since I only shop once per week, the 24 hour limit would exclude this data.

I am confident that in my initial 2012 scoring I did not limit my response to events of the past 24 hours. I suspect I filled it out in a manner similar to the 2013 update; I based my response on remembered events from the past month or two.

This raises another concern. The instrument relies on accurate reporting by the subject. I have learned from my work with Dr. H (not Doctor H2 – that is someone else) that I have problems with self awareness. So it is reasonable to expect Recency bias, Availability heuristic, and other subjective distortions of the data.

Despite these concerns, I find the instrument to be of value. Apart from one, or two, items, it is readily accessible to a lay person and appears to cover the full scope of what I experience as problems. I cannot think of any element I would wish to add (I need to go back and compare the inventory with some of the details in Dr H’s reports. She may have identified other issues. Her reports are difficult for me to digest as they contain medical language unfamiliar to me. ).

Subsequent to locating the September 2012 Inventory, I created a new spreadsheet containing the initial 2012 data and entered my August 2013 responses. I used the spreadsheet’s internal functions to tabulate totals and avoid the introduction of mathematical error.

The table below provides my original 2012 scores, the corrected 2012 scores (corrected due to the fact that the original tabulation was performed manually and was found to contain errors. The Corrected 2012 column accurately reflects accurate 2012 totals) and my 2013 scores for each section.

I have revised the data in the September 2012 Inventory post so that it now reflects the original data recorded in September of last year. I went through all elements of the post and edited it to accurately reflect the original document. If I only entered a single “?” in the original, that is what now shows in the post. If the column totals are incorrect that also reflects the original document.

The table above provides the “as entered” section totals for 2012, the corrected  section totals for 2012 (with the calculations being performed by software functions rather than by hand) and section totals for 2013.

It appears from these totals that I am getting worse. This is an incorrect conclusion. The increase in the 2013 totals derives from entering data for items that I failed to score in 2012. I do not know the reason these were omitted in the original. The second reason for the increase has to do with my increased level of self awareness. This has resulted in significant jumps in my scores for items on despair and depression. I believe I am improving but the item scores do not reflect this fact.

I am still doing a workup of the different items and trying to determine how best to present the information without spending the next month posting on this single topic. I am trying to develop a synopsis of the major variances between 2012 and 2013 and this should be available in the next post.