H is for... Humping

Good morning, boys and girls! Today's blog post concerns the incidence of humping on the production oceans.

Chat logs from twenty visits to each of Cobalt, Midnight, Hunter, Opal, Sage and Viridian oceans were analyzed for the occurrence of humping. Visits were made by three pirates: Apollo, Sun and Zeus. The pirate logged in was recorded along with the ocean and evidence of humping activity.

Analysis of Humpitude by Ocean indicated that Midnight was most likely to hump (mean 0.55), followed by Viridian (0.40), Sage (0.20), Cobalt (0.10), Hunter (0.05) and Opal (0.00). Incidence of humping was not signficantly different between Midnight and Viridian (p=0.344). However, incidence of humping was significantly higher on Midnight and Viridian than on Sage, Cobalt, Hunter and Opal (p<0.05 in all cases).

Analysis by Pirate indicated that Pirate was a confounding variable in the analysis, with Apollo significantly more likely to be humped than either his father or his sunny alter ego (F=23.719, p<0.001); the variable Pirate was therefore included as a covariate in further analysis.

Comparison of means by Ocean and Pirate showed that Apollo was humped on all oceans but Opal, with Midnight being the most frequent (0.89). Sun was humped on Midnight, Sage and Viridian, while Zeus was only humpitudinally advantaged on Midnight and Viridian. A univariate analysis of variance on dependent variable Humpitude indicated that both Ocean (df(5), F=4.459, p<0.001) and the covariate Pirate (df(2), F=12.364, p<0.001) were significant in the prediction of Humpitude. The interaction Ocean x Pirate was not significant (p=0.734), and a parsimonious model excluding this interaction is recommended. Adjusted R Squared was equal to 0.371, indicating that 37.1% of the variance on the dependent variable Humpitude could be explained by the two variables Ocean and Pirate. Analysis of residuals continues.

Half of this analysis was carried out in SAS brand statistical analysis package, and half was carried out in Matlab brand statistical analysis package. Matlab performed the analysis 25% faster, and removed more stubborn stains.

1. Times of chat logs were noted, and analysis by game time may also yield interesting information.
2. The main research question outstanding is to discover what exactly is meant by humping. The researcher would feel much better if he knew whether it was a good thing or a bad thing, and proposes some qualitative research and analysis to investigate the phenomenon of Humpitude in more detail, using focus groups and in-depth interviews, followed by textual analysis and Critical Incidents Techniques on the resulting transcripts.
3. The parallel research project on licking is currently delayed, as the researcher's stats packages asploded when he tried to import the subscription ocean lickage data.
4. Now wash your hands.

H is also for... hats, hugs and heroes, all of which were part-drafted in my H for head. H for however, H for humping just won out in the end!