Marmalade Quill Shelf Institute

Tag: statistics

  • An Open Letter To Everyone Who Says It Was Just The Wind

    An Open Letter To Everyone Who Says It Was Just The Wind

    We write to address a persistent claim: that a fallen object was moved by the wind.

    We have examined this claim in depth. We have measured the wind. We have measured the cat. We find that the cat is stronger.

    Evidence Against The Wind

    • The wind has never been seen holding a paw out.
    • The wind does not look at you afterwards.
    • The wind does not purr.

    Evidence For The Cat

    The cat was present. The cat was calm. The cat was already looking at the object when the window was still closed. We consider this decisive.

    Methodology

    We placed a lightweight cup on a table by an open window and recorded wind speed throughout the day. The wind peaked at midday, at which point nothing happened. At 12:03 a cat entered the room and the cup fell. The Laboratory finds this statistically persuasive.

    A Note On Curtains

    We acknowledge that curtains sometimes move in the wind. But a curtain that moves is not a curtain that pushes. A curtain, we remind readers, is merely a flag of surrender.

    Reply From The Wind

    No reply has been received. We take this as an admission.

    Appendix A

    The Laboratory has also received a letter from a rival group, who claim that draughts explain 12 percent of falls. We have replied that the remaining 88 percent are explained by the cat, and that the 12 percent were probably the cat as well, in a hurry.

    Appendix B

    For completeness we have measured the speed of the wind against the speed of a cat who has just seen a bird. The cat won by a wide margin and then lost interest.

    Conclusion

    We rest our case on the table, gently, near the edge.

    We invite the wind to submit a rebuttal. It may leave it on a table edge.

  • Statistical Appendix: How We Measured Disappointment

    Statistical Appendix: How We Measured Disappointment

    The Laboratory developed the Disappointment Index to measure how let down a cat appears when an object does not fall as expected. It uses four categories.

    • Zero: the object fell. The cat was pleased.
    • One: the object fell, but silently. The cat was mildly cross.
    • Two: the object did not fall. The cat looked at the object for a long time.
    • Three: the object did not fall, and the human put it back. The cat looked at the human for much longer.

    Validation

    The Index has been validated against the rate of tail twitching, which correlates strongly with disappointment and also with lunch. We have controlled for lunch by timing the trials before breakfast, which has made everybody sad.

    Use In Industry

    Several manufacturers of mugs have asked to license the Index. We have declined, since it is not for sale, and since they could not be trusted to hold a cat steady.

    Limitations Of The Index

    The Index has limits. Cats are complex creatures, and a single twitch may reflect disappointment, joy or an itch. In the early version of the Index, a cat scratching an ear was recorded as four on the scale, which caused unnecessary worry.

    Future Work

    We plan a Version 2 of the Index that includes a fifth category, Resignation, to be applied to cats who have given up on the whole thing and gone to sleep. A sixth category, Smugness, is already in draft.

    Case Example

    In one trial, a cat named Biscuit watched a cup remain stubbornly on a table for forty minutes. She scored a three. When the cup fell, she scored a minus one, which we had to invent on the spot and which now sits in an annex.

    The appendix is available in print on request, in a font large enough for a cat to read from a distance.

  • A Study Of Why Pens Roll Away From You

    A Study Of Why Pens Roll Away From You

    Study 98 investigated a common complaint: that a pen, when dropped, rolls away from the person who dropped it. We assembled nine pens and a desk and set out to find out why.

    Findings

    All nine pens rolled away from their owner. Seven rolled beneath furniture. Two rolled onto the cat, who treated them as gifts. We conclude that a pen has a clear opinion of the person holding it, and we have not found a way to change that opinion.

    Theories Considered

    • The desk is tilted. It is not.
    • The pen is round. It is, and this was noted.
    • The pen dislikes you. This hypothesis remains open.

    Conclusion

    Method

    Pens were selected at random from the Laboratory stationery cupboard. Each was placed at the head of a wooden slope, released, and tracked by a camera and a very focused cat. The cat tracked the pens by following them. The camera tracked the cat.

    Data

    Rolling distance averaged 61 centimetres, with a standard deviation of 12. The outlier, a pen of unusual width, rolled for 3 metres and vanished beneath a filing cabinet. We recorded it as a successful escape.

    Practical Advice

    If you must keep a pen on a desk, place it in a cup. If you wish to keep it in a cup, tie the cup. If you wish to keep the cup, have a stern word with the cat.

    Further Study

    Study 99 will examine why a pencil, when dropped, rolls less but sounds smug. The Laboratory has already budgeted two biscuits for the result.

    Reviewer Comment

    A reviewer asked whether the pens might simply be shy. The Laboratory has added shyness to the list of variables, between humidity and mood.

    A pen, once dropped, becomes a small independent creature with its own travel plans. The Laboratory recommends tying it to the desk, or replacing it with a pencil, which is flatter and less proud.

  • Peer Review Report: ‘On the Sound Of A Fallen Fork’

    Peer Review Report: ‘On the Sound Of A Fallen Fork’

    The Laboratory was recently asked to review a paper titled “On the Sound of a Fallen Fork” from the Department of Crashes at a neighbouring institute. We reproduce our report below, with their names removed and our feelings preserved.

    Summary Of The Submission

    The authors claim that a fork, dropped onto a tile floor, produces a sound that varies according to the mood of the person who dropped it. The authors present four graphs, two photographs and a sentence that appears twice.

    General Comments

    The topic is of interest. The execution is not. The paper confuses a fork with a spoon on page three, then corrects itself on page five, then confuses them again on page six. A reader may begin to suspect that the authors have only seen cutlery on television.

    Specific Concerns

    • The sample size is one fork. We recommend at least two, and ideally a whole drawer.
    • No cat was present at any of the trials. The authors do not explain how the fork fell without help.
    • The word “clang” is used 31 times. The word “clink” is used 4 times. Neither is defined.
    • Figure 2 is a drawing of a fork, and appears to have been done by a child. We hope it was a talented one.

    Statistical Analysis

    The authors state that their results are significant at the 5 percent level. Our statistician, a cat who prefers not to be named, looked at the figures for several minutes and then yawned. The Laboratory interprets this as a rejection.

    Recommendation

    We recommend that the paper be rejected, then accepted, then rejected again, in order to give the authors a sense of the full experience. We also recommend that they add a cat.

    Detailed Remarks On Methodology

    The authors describe their method as rigorous, which is generous. A fork was dropped from the height of a human elbow onto a surface described only as hard. There is no record of the temperature of the room, the age of the fork, or whether it was in a good mood. In our experience a fork in a bad mood sounds noticeably sharper.

    The authors also appear to have used the same fork for all trials. A fork subjected to repeated falls will develop a certain world weariness, and this will bias the results. We recommend a fresh fork for each trial, and a moment of silence for the retired ones.

    Remarks On The Literature

    The reference list contains 14 items, of which three are the authors’ own previous papers and two are recipes. We checked the citations. Reference 9 does not exist. Reference 11 exists but argues the opposite. Reference 12 is a photograph of a cat, and we are happy to say it is the strongest source in the paper.

    Remarks On Ethics

    The authors do not state whether the fork consented. We would like to see an ethics statement, a risk assessment and a small cushion.

    Questions For The Authors

    • What was the fork thinking?
    • Why was the floor tile and not wood?
    • Why does the second graph show the sound getting quieter with time, when the sound clearly continues in the reader’s head?
    • Is the fork a member of the authors’ family?

    Confidential Comments To The Editor

    This is the best paper we have read this week. We recommend rejecting it solely so that we may submit something similar next month.

    Reviewer Conduct

    This review was prepared by three cats, one of whom did the reading, one of whom did the writing and one of whom lay across the printout throughout. We believe that is a standard division of labour.

    We offer these comments in the spirit of collegiality. We will be at the conference, near the buffet, standing very still, and watching.

  • Frequently Asked Questions About Our Height Standard

    Frequently Asked Questions About Our Height Standard

    We receive many questions about the Laboratory’s official drop height. Here are the ones we find we can answer.

    Why 72 centimetres?

    The standard height was set in the Laboratory’s founding year by Professor Quill, who sat on a stool, extended one paw, and said “about there”. Gravel then measured the distance with a ruler and found it to be 72. Nothing about the figure is scientific. It is, however, consistent, which is more than we can say for the stool.

    What happens at 73?

    Nothing, according to the data. At 73 centimetres the object falls in exactly the same way, with the same sound. We maintain the standard because a standard is a thing that cannot be argued with, and arguing is the Laboratory’s second favourite activity.

    Can we test other heights?

    Yes, on request. Our Variable Height Programme offers heights from a low 20 centimetres up to a dizzying 140, which requires a ladder and a lot of paperwork. Please note that above 100 centimetres the cat must sign a waiver, which is a paw print on a napkin.

    Does the object matter?

    The object matters greatly. A pen falls with a clean little tick. A glass falls with a bright splash. A cushion falls with a sigh and is not covered by our standard. We therefore classify objects by mood.

    Is it cruel to the objects?

    The Laboratory takes the welfare of every object seriously. Objects are provided with a soft landing mat when requested, and are never subjected to more than three drops in a day. Objects who wish to lodge a complaint may do so by remaining on the floor.

    Can I visit?

    Public tours run on Wednesday mornings. Please do not bring unlabelled glassware. It will be tested.

    Do you accept donations?

    Why Not 70 Or 75?

    Several colleagues have suggested rounder numbers. The Professor replied that a round number is the work of a lazy mind, and that 72 is a number with character. It is a dozen, times six. It is also, as Gravel points out, the exact height of the lab kettle, which makes it easy to remember and quick to boil.

    How We Verify The Height

    Each morning a technician lies on the bench and holds a ruler against the floor. A second technician lies on the floor and holds a ruler against the bench. The two rulers are compared and disagree by about four millimetres. This is recorded as measurement error and then, quietly, as a character flaw.

    Does Height Affect The Cat?

    A small number of cats become more serious at greater heights. At 72 centimetres a cat is relaxed. At 100 she becomes thoughtful. At 140 she goes very still and says nothing, which our team interprets as respect or as a plan.

    Can I Choose My Own Height?

    You may. The Laboratory will happily test your preferred height, though we charge a modest fee for the privilege and an additional fee for the ladder. We will not test heights measured in cats, since the cats object.

    Where Can I Read More?

    Our height standard is documented in the Laboratory Handbook, a leaflet of two pages, one of which is a drawing of a stool. A copy is available at reception, if the receptionist has not sat on it.

    A Final Word

    Please do not send us questions about height in imperial units. The Laboratory does not recognise inches, as no cat has ever been able to find one. Our measuring staff also object to the word foot, on the grounds that it is rude.

    We do. Our most useful gifts are small ceramic items, cheap mugs, and any pen that has not been missed. Please wrap them carefully. Then please tell no one.

  • Results: Does a Vase Fall Faster When Watched?

    Results: Does a Vase Fall Faster When Watched?

    The Purrfect Drop Test Laboratory is pleased to publish the results of Study 114: Does a vase fall faster when watched? Our principal investigator, Professor Dame Marmalade Quill, has long suspected that observation changes the speed of a descent. She has been proven right, though not in the way she expected.

    Abstract

    Forty vases were placed at the edge of a standard laboratory bench. Twenty were observed by a trained cat. Twenty were left alone. The vases were then pushed, in a manner consistent with scientific protocol, by a second cat, whose identity we protect for ethical reasons and because she bites.

    Method

    Each vase was weighed, named and given a number. The names were chosen by the laboratory technician, a thoughtful grey cat called Gravel. The vases that were observed were looked at continuously from a distance of one metre. The vases that were not observed were looked at, briefly, from a distance of two metres, which the team agreed was close enough to count as not looking.

    • Sample size: 40 vases, 2 cats, 1 technician
    • Drop height: 72 centimetres, measured by a cat lying down
    • Timing: a stopwatch, held by someone who was not paying attention
    • Surface: laminate, with a small scratch that has become a team mascot

    Results

    Observed vases fell, on average, 0.02 seconds more slowly than unobserved vases. The difference is small, but our statistician describes it as significant, and our statistician is a cat who sits on the calculator. We therefore accept the result.

    A more interesting finding concerned the observer. In 17 of 20 observed trials the observing cat turned away at the moment of impact. Professor Quill refers to this as the Averted Gaze Effect and believes it demonstrates a deep, unspoken moral discomfort. Gravel believes it demonstrates that the cat had seen enough vases.

    Discussion

    The simplest explanation for the delay is that a watched vase is a vase with something to prove. It hesitates. It reconsiders. It falls anyway, but with a certain poise. The unobserved vases, by contrast, tumbled with the careless speed of an object that knows nobody cares.

    “A watched vase never falls, until it does. Then it makes a lovely noise.” Professor Dame Marmalade Quill, in the lab tea room.

    A second possibility is that the observing cat blinked slowly, which is known to slow time. We will investigate this in Study 115, which has been funded by a grant of one tin of tuna.

    Limitations

    The study has several limitations. First, the stopwatch was held by a person who was also eating a sandwich. Second, three vases were claimed by the cat as personal property before the trial and had to be excluded. Third, the observer cat fell asleep during trial 29, and we recorded the vase as having been observed in spirit.

    We remain, as ever, deeply confident. All results are available in a folder on a shelf, which may be reached by climbing. The Laboratory thanks all participants. The vases were not available for comment.

    Next Steps

    Ethics Statement

    All vases were treated with appropriate respect. None were harmed beyond the extent required by the study, which in most cases was entirely. Surviving fragments have been returned to the shelf marked Evidence, where they are visited by the Professor each Friday.

    Acknowledgements

    The authors thank Gravel for sweeping, the second cat for pushing, and the first cat for watching. We also thank the tea trolley, without which the team would have been much shorter tempered, and the radiator, beneath which a great deal of data was later found.

    Study 115 will examine whether a vase falls faster when sung to. Study 116 will test whether it falls faster when insulted. The Laboratory is also discussing a study on whether it falls faster when nobody is in the room at all, but nobody can agree who would watch.