Procedure for Computing Relative Importance Ranking of All Articles
GMi =  (YSi × SSi × CSi) | eq 1: Geometric mean |
| Fwij = Fij × GMi | eq 2: Consolidated weighted score |
| Rij = Rank(Fwij, Fw11: Fwnk) | eq 3: Rank including Ties |
| OP = {Sum[If (Fwij < Fw11: Fwnk, 1, 0)]} | eq 4: Ordinal position |
| TC = Countif ⌊Rij ∈ (R10: Ri(j−1))⌋ | eq 5: Tie count (R10 is an empty cell) |
| Rij = OP + TC + 1 | eq 6: Rank correction to account for Ties |
| RRij = (n − Rij) + 1 | eq 7: Rank in reverse order |
| RankCj = Rank(Cj, C1: Ck) | eq 8: Cj = ∑ RRij where i = 1 … n and j = 1 … k |
Notes: Equation 1 was used to determine the geometric mean (central tendency) of the weighted scores for each article.
In equation 2 the weighted scores were multiplied by the geometric mean to determine a single Consolidated Weighted Score (CWS) for each article.
Equation 3 was used to compute relative importance ranking of each of the CWS relative to the entire population of CWSs.
Equation 4 was used to find the ordinal position of the CWSs. Some ordinal positions had the same value, which are termed “Ties.”
Equations 5 and 6 were used to compute the ordinal position of the CWS taking into account the ties.
Equation 7 was used to perform the ranking in reverse order.
Equation 8 was used to calculate a single consolidated ranked score for each transformation trigger (by adding columns of the rank matrix). The larger the consolidated ranked score the higher the relative ranking of the transformation trigger.