The Mean Delay I experienced during these four years of "training" was

    The Total Delay experienced on the route Kalmthout-Leuven-Kalmthout during this period was

    A complete breakdown of the variation in this delays by means of an analysis of variance (ANOVA) might reveal some suggestive explanations for the delays experienced at different periods... Variables looked at to split up the delays were the period of the delay (PERIOD, 4 levels) as reflected in the different subpages of this web content, the month in which the delays occurred (MONTH, 12 levels), and the day of the week (DAY, 5 levels).
    These three variables and their interactions managed to explain 31.4% of all the differences in the the delays experienced during these four years!! Beyond DAY, that showed a minor overall effect only, it were primarily MONTH and PERIOD that added to this model.

    Although not significant, the effect of  DAY, F(4,676) = 1.87 (ns), reflected a clear trend. On average, the largest delays were experienced in the beginning of the week, on mondays and tuesdays, as shown in the next figure. Wednesdays clearly had the lowest delays. Finally, delays on thursdays and fridays very much reflected the overall mean delay of about 14 minutes. What's happening here? Does the weekend, which is not part of these analyses since I never take a train in the weekend, play a role? Is the railroad staff still tired of the weekend on mondays and tuesdays, and already looking forward to the weekend on thursdays and fridays? Or, is there simply more passengers on these 'delayed days' than on wednesday? Who tells me?

    Regarding MONTH, a highly significant trend shows up, F(11,676) = 7.67 (p < .001), which on its own explains 9% of all the variation in the delays experienced. The next figure clearly shows what's the meaning of this effect... Wintertime clearly plays a role! Massive delays from September on, peaking in November and December. Strangely, January does not confirm this general trend... Good intentions of the railroad staff at the beginning of the year, resulting in minor delays, on its turn producing satisfied passengers, this way anticipating to the traditional increase of the prices of the tickets at the beginning of February? The summer holidays, July and August, clearly show the lowest delays. Less passengers, and more importantly, less trains on the rails, that's the reason!

    The results of PERIOD, F(3,676) = 6.10 (p < .001), kind of meant a support for the changes made to the timetables by the railroad managers. On average, delays decreased over the four subsequent timetables of periods that were studied. See in the next figure... Regarding this last effect, it must be stressed that all effects in this ANOVA are unique effects. This means that the differences between periods shown in the figure below, occurred above and over the previously reported differences by month. So, the low delays in the final period really do reflect an improvement and not simply summertime!

    Finally, none of the interactions between PERIOD, MONTH and DAY reached significance, although we expected some (e.g., the interaction between PERIOD and DAY, reflecting different day patterns for the various periods).

    In sum, simply looking at PERIOD, MONTH and DAY clearly explains lots of the delays that occurred during the last year. This is an important finding that needs subsequent research. What concrete variables are behind these simple things? What other variables might be of interest? Although already reflected in the MONTH variable, the weather might be important. Another variable of interest, yet hard to include in the study, is the number of passengers at each train.

    Remember, the new hours during the last period again took longer than the old ones. We kept on getting slower all the time ! The next picture nicely summarizes what I mean by this... It represents the duration of my traveling as it should have been (as printed in the official timetables)... I started my commuting between home and office in 1997 at a good regime, that is, 750 minutes a week, however with very much delays (see before). Since the introduction of the IC/IR98 project, we simply became increasingly slower! First a massive increase to 850 minutes a week, then slightly going upwards to 885, and finally 905 minutes a week!
 

    And now, since November 2001, for the first time, we became faster! And how! The route Kalmthout-Leuven-Kalmthout now takes exactly 820 minutes a week! That's 85 minutes a week better, if everything runs well... So, a silver medal overall, and a golden one since the introduction of IC/IR98 !!!

    How did this happen? Well, look a the new timetable:
 

    So, we gain minutes in the morning! Every day I can sleep 17 minutes longer! How did this change happen? Well, that's a very personal story...

    What was the problem? The last change made to the timetable, early June 2001, involved the morning train leaving the departure station earlier (see here for more info), again. This again had to do with changed timetables for another, more important train, the international train Amsterdam - Brussels, that is using the same tracks. As a consequence, our morning train, which is mentioned as IR (inter-regio) in the books, now started to run slower than a local train on the same route!!!

    The day after this surrealistic change was made to the official timetable, that is, June 11th 2001, I immediately contacted the regional railroad manager at the station of Antwerpen-Berchem, to complain on the new situation, and I immediately presented some of my main arguments for a solution to this situation. He answered me with a kind off an umbrella strategy, that is typical of the NMBS. "This change of the timetables was not his responsibility, and I should contact someone else..." But I insisted, and he gave me a name of an important chief at the headquarters of the NMBS, Mr. Freddy Pardaens. He was mentioned to as the one responsible for the changes in the timetables...
    So, arrived at work, I immediately called Mr. Pardaens by telephone and explained him the problem. First, he tried to ignore my remarks, as being the useless comments of simply another frustrated traveler... However, again I insisted and presented him all my arguments - that a surrealistic situation has emerged namely a IR-train running slower than a train, that we just were bearing the consequences of a change in the timetables of an international train, that giving the early hour, transported barely no passengers, that our semi direct interregional train followed the regime of a local train, etc. - and then he provided me a reason for the latter. "Following the general rules of the NMBS, all first trains of the daily regime, starting at 6 a.m., should run as a local train, whatever their status is. Of course, international trains fall out of this general rule". Aha... that was a clear rule, and indeed our morning train was such a first train in the daily regime...
    WAS! Indeed, WAS!! Because since the last change made at June 10, our train in fact became the last train of the nightly regime! It now left Essen, the starting station, well before 6 a.m.!! After I presented this, I agree silly, argument to Mr. Pardaens, he was prepared to listen to me, and suggested to write down all my arguments and send them to him. I agreed and so I did.
    Some weeks later, after he studied my letter, Mr. Pardaens himself contacted me, as I requested to give me a follow up to my remarks. He told that he managed to convince his superiors that this really was a problem, and kind of thanked me that I pointed them to this problem. "They themselves over there in the central offices in Brussels were not able to overlook all the consequences of changes made to the timetables at once, and depended on the information of lower railroad workers. However, these latter part of the staff didn't supply all of the necessary information." (Mr. Pardaens' words).
    And yes, they agreed to take a look at my solution, that is, changing the regime of the early morning train to a real IR-regime, rather than a local one, so this train no longer would interfere with the international train from Amsterdam, and in the same time, would run to Antwerp in a shorter time. Mr. Pardaens went on to say that they would do an examination of the consequences of this change of regime, consequences for the travelers on our line, because the early morning train now no longer would stop a some stations it halted before. We agreed on this, and I was promised to receive the results of that research.
    Some weeks later I saw the ticket collectors on the early morning train counting the passengers and asking who entered where to go where... They of course didn't know why they were doing that, but I did and of course did not tell them. They kind of looked frustrated with the extra work.
    Weeks passed on, and I did not hear anything from Mr. Pardaens... Did he forget me at last? Was the case closed due to the outcomes of the research? But, then, no longer expected, on October 18th, about four months after my initial question, I received an e-mail from Mr. Pardaens (yeah, they have computers at the NMBS!) stating:

"Geachte,
Gevolg onze vorige telefonische contacten.
Vanaf 5 november e.k. rijdt deze Ir volgens kadans.(V. Essen 6u16).
Hoogachtend,
F. Pardaens"

Or, in English:

"Dear,
Following our earlier telephonic contacts.
From November 5th upcoming, this train runs the IR regime. (Starting in Essen at 6.16 a.m.).
Sincerely,
F. Pardaens"

    Immediately understanding the content of this relatively mysterious mail, I promptly called Mr. Pardaens by phone, and he confirmed what I thought. From November on, the early morning train would run as a real interregional train, leaving Essen about 17 minutes later, no longer interfering with the international train, simply running faster!

    As I suggested him before! I myself managed to change the NMBS timetable!! Indeed, I'm a bit proud!! Indeed, I sleep 17 minutes longer every morning since November! And happy I am.

    With this happy message I want to end: When you experience a concrete problem using the train, analyze it clearly, try to contact the right man or woman at the right place, high enough in the hierarchy of the railway company, and overshoot him or her with all your arguments, preferably formulated in a bureaucratic formal language, because that's what these persons know best. Keep insisting and ask for immediate feedback! This really helps!

    This is the end of my railway related pages. I finish this research project here! It's been enough! Moreover, from now on, I only will take the train three days a week, because I'm working home part-time. This would make my data only a good estimation of delay reality, rather than a true description of it. Hope you enjoyed reading. As mentioned before, all comments welcome at here.
 

p.s. And indeed, these locs finally arrived, but are still not all in use, because when I arrive in Leuven in the morning, I see them lined up in rows of six, waiting to be used... In the mean time, NMBS is undergoing major organizational changes. We'll see what comes of this... More strikes??? Probably...