Skip to content
Open navigation
OpenOpen workspace

Worked exampleHospitality

Were bookings canceled more often at the city hotel or the resort?

Bookings that ended canceled, by property: Resort Hotel: 10,831 of 40,060 (27.0%). City Hotel: 32,186 of 79,330 (40.6%). The share was 13.5 percentage points lower for Resort Hotel.

The table

Final reservation status by property. Percentages use each property’s bookings as the base.
Final booking status \ PropertyResort HotelCity HotelTotal
Canceled10,83127.0%32,18640.6%43,017
No-show2910.7%9161.2%1,207
Checked out28,93872.2%46,22858.3%75,166
Total40,06079,330119,390

Count / Column %. Each column total is the base for that column’s percentages.

Shaded cells are the two shares the answer compares.

Canceled bookings by arrival year

The table above pools every arrival year. Here its Canceled row is split by the year each guest was due to arrive. Each share uses that hotel’s bookings for that year as its base, so every percentage can be checked against its counts.

  1. 2015

    July–December 2015, 6 of 12 months

    Incomplete year
    2015: canceled bookings and base by property
    PropertyCanceled / bookingsCanceled share
    City Hotel5,882 of 13,68243.0%
    Resort Hotel2,069 of 8,31424.9%

    Gap: the City Hotel share was 18.1 percentage points higher than the Resort Hotel share.

  2. 2016

    January–December 2016, all 12 months

    2016: canceled bookings and base by property
    PropertyCanceled / bookingsCanceled share
    City Hotel14,852 of 38,14038.9%
    Resort Hotel4,817 of 18,56725.9%

    Gap: the City Hotel share was 13.0 percentage points higher than the Resort Hotel share.

  3. 2017

    January–August 2017, 8 of 12 months

    Incomplete year
    2017: canceled bookings and base by property
    PropertyCanceled / bookingsCanceled share
    City Hotel11,452 of 27,50841.6%
    Resort Hotel3,945 of 13,17929.9%

    Gap: the City Hotel share was 11.7 percentage points higher than the Resort Hotel share.

The three years add back to the table above: 32,186 of 79,330 City Hotel bookings and 10,831 of 40,060 Resort Hotel bookings were canceled.

Gaps are computed from unrounded shares, so they can differ by 0.1 from subtracting the rounded figures.

What this split can and cannot tell you

  • This is a descriptive split by scheduled arrival year, not by the date a booking was made or canceled.
  • 2015 covers only arrivals from July to December, and 2017 only from January to August. These edge years miss different seasons, so they are not like-for-like with the full year 2016.
  • Comparing shares across years or hotels describes what was recorded; it does not explain why the rates differ. The hotels differ in location, guest mix and booking channels, and none of that is held constant here.

Original source datasetRun a crosstab on your own data

Check the reading before you quote it

Pick the statement this table supports, then open each option to see what the counts say. Every figure here is computed from the table above.

Does the cancellation gap between the two hotels hold in each arrival year?

  1. Yes. The City Hotel share was higher in every arrival year.

    Supported. Over all arrivals, 40.6% of City Hotel bookings and 27.0% of Resort Hotel bookings were canceled. Split by arrival year, the City Hotel share was higher by 18.1, 13.0 and 11.7 percentage points in 2015, 2016 and 2017.

  2. The smaller gap in 2017 shows the two hotels converging.

    Not supported. 2017 covers arrivals only from January to August, and 2015 only from July to December. These edge years miss different seasons, so a smaller gap there is not like-for-like with the full year 2016.

  3. The table shows that a city location causes more cancellations.

    Not supported. The hotels differ in location, guest mix and booking channels, and none of that is held constant here. The table describes recorded bookings; it does not explain them.

The figures behind the check

  • Arrival coverage: July 2015 to August 2017; only 2016 has all twelve months.
  • Canceled share over all arrivals: City Hotel 40.6%, Resort Hotel 27.0%.
  • Gap by arrival year, City Hotel minus Resort Hotel: 2015 18.1, 2016 13.0, 2017 11.7 percentage points.

What to carry forward

Quote the direction with its arrival coverage, July 2015 to August 2017, and label 2015 and 2017 as partial years whenever yearly gaps are compared.

Reproduce the exact tableOpens the calculator with this table’s counts, labels and category order, so each figure can be checked.

Use this result

  1. Reproduce it

    Open these exact counts in the chi-square calculator. It loads the same table, with the same category order, and computes the percentages, χ² and Cramér’s V in your browser. Change a cell to see how the result moves.

  2. Cite it

    Copy this reference. The data credit and licence below travel with it.

    Were bookings canceled more often at the city hotel or the resort? — Crosstabs. https://www.crosstabs.com/examples/ht01-how-did-the-mix-of-canceled-no-show-and-checked-out-bookings-differ-between-the-two-hotels

    Source

    Scope
    One Algarve resort hotel and one Lisbon city hotel, Portugal.
    Period
    Scheduled arrivals 1 July 2015 – 31 August 2017
    One record is
    one reservation, including canceled bookings

    Data: Nuno Antonio, Ana de Almeida and Luís Nunes (2019). Hotel booking demand datasets. Combined CSV distributed by Jesse Mostipak on Kaggle.. Licensed under CC BY 4.0. Crosstabs selected, recoded and aggregated the records into this table.

    Download these counts: CSVJSON

  3. Run it on your file

    Open a CSV, Excel or SPSS file to build the same kind of table. The file is read in this browser tab: nothing is uploaded and no account is needed.

    Run a crosstab on your own data

Why ask this

Revenue managers set overbooking and deposit rules per property. Cancellations and no-shows cost a hotel in different ways, so the final booking status has to be split into all three outcomes before the properties are compared.

How to read it

Read down each property’s column. No-shows are a separate row from cancellations: 0.7% of resort bookings and 1.2% of city-hotel bookings, small next to the canceled share.

Column percentages answer “within each column, how often does each row occur?”. Switch to row percentages and the question turns around: “of the records in this row, how were they split across the columns?”. Both are correct; they answer different questions, so say which one you used.

What it means

A larger share of city-hotel bookings ended canceled. The two properties differ in location, guest mix and booking channels, so this describes the two hotels; it does not isolate a city-versus-resort effect.

Cramér’s V = 0.137. It summarizes the whole table, all 3 × 2 categories, on a scale from 0 (no association) to 1. It does not say which cells differ, and it is not a measure of impact.

Comparing only Resort Hotel and City Hotel, and only Canceled against every other row, Cramér’s V = 0.133. That number describes this one contrast, not the whole table.

Limits

Booking records from two properties only. They do not represent Portuguese hotel guests or the industry as a whole.

Keep exploring