By default, ANY-maze reports distances in metres and speeds in metres per second. However, if you are working with a small animal, such as a zebra fish larva, then having everything reported in metres is not very convenient. In this case you could alter the units, perhaps to millimetres.

You can change the units at any time except while a test is actually running; ANY-maze will automatically update all reports to use the new units.

Changes affect three things: Distances, speeds and distance integrals (such as the CIPL measure).

Setting the level of significance for statistical tests

ANY-maze includes the ability to perform a number of statistical tests (ANOVA etc.) which you can use to assess the results of your experiments. The level of significance (alpha) in these tests is an analysis option; it's set to p ≤ 0.05 in all new protocols but you can change it to any value between 0.01 and 0.50.

Choosing how to manage latency of events which don't occur

ANY-maze can measure the latency of a number of events, such as 'Latency to the first entry to a zone', 'Latency to a key press', and 'Latency to the start of a sequence'.

Of course, there always exists the possibility that a particular event doesn't actually occur in a test - the animal doesn't enter the zone, the key isn't pressed, or the animal never performs the sequence. In these cases, the latency value will usually be reported by ANY-maze as undefined.

Considering a latency to be undefined is strictly correct (as we have no data for it), but it means that the Ns for the latency measure will be reduced, as ANY-maze considers an undefined result as a result which simply doesn't exist.

For example, imagine you tested 3 groups of 10 animals in the plusmaze. During the tests, all the animals entered the open arms at some point, except for two animals in the first group which spent the entire test in the closed arms or central area. If you then analysed the 'Latency to first entry to the open arms', you would find that the N for this measure for Group 1 would be 8, rather than 10, because for two of the animals, ANY-maze wouldn't have a result at all.

To prevent this reduction in groups' Ns, some researchers choose to consider the latency for events which don't occur to be the test's duration rather than being undefined. If you want to, you can set ANY-maze to work in this way by checking the box labelled Use the test duration as the latency for events which don't occur.

Choosing how to manage undefined averages

Some of the measures that ANY-maze reports are averages, for example, 'Average duration of visits to a zone' or 'Average time to complete a sequence'. Clearly, these values will be calculated by dividing one value by another, for example the total time in a zone by the number of visits to the zone. This however, can cause problems if the denominator is zero - for example, the animal never entered the zone or never completed the sequence.

In these cases, ANY-maze will normally consider the result for the average to be undefined, but as an undefined result is treated as a result which simply doesn't exist, this will cause the group Ns for the relevant measure to be reduced.

For example, imagine you tested 3 groups of 10 animals in the plusmaze. During the tests, all the animal entered the open arms at some point, except for two animals in the first group which spent the entire test in the closed arms or central area. If you then analysed the 'Average duration of visits to the open arms', you would find that the N for this measure for Group 1 would be 8, rather than 10, because for two of the animals, ANY-maze wouldn't have a result at all.

To avoid this reduction in the groups' Ns, you can tell ANY-maze to consider the result of an average for which the denominator is zero to be zero rather than undefined. To do this, simply check the box labelled Use zero as the result for undefined averages.

Choosing how to manage undefined path efficiency to a zone

A similar situation to that described above (for undefined averages), exists for the path efficiency measure, when measured as the efficiency to the first entry to the zone. This is calculated as the straight line distance between the first and last position, divided by the total distance travelled between those positions. Obviously if the animal has never entered the zone, the path efficiency can't be calculated - again, the N would be reduced by the number of animals who never entered the zone. You can avoid this by telling ANY-maze to use zero as the path efficiency (thus ensuring that N matches the number of animals) by checking the box labelled Use zero as the result for undefined path efficiency to zones.

Choosing how to calculate a zone's CIPL

The Corrected Integrated Path Length (CIPL) is usually calculated based on the animal's path and average speed until it first enters the zone. If the animal never enters the zone then this average speed will be undefined and as a consequence the CIPL will be reported as #N/A.

You can change this behaviour by choosing the option to Calculate CIPL for a zone based on the animal's path throughout the test, in which case, in those tests where the animal does not enter the zone, the CIPL will be based on the animal's average speed throughout the test and therefore a value will be reported.

Switching this option on also affects the CIPL result for tests where the animal enters the zone more than once, as when switched on it causes the CIPL to be based on the animal's path throughout the test and not just until it first enters the zone.

For example, if in a water-maze experiment the animal swam over the platform, but the test continued, and only later did it stop on the platform (causing the test to end), then with this option switched off, the CIPL would be based on the average speed and path until the first platform entry, whereas with this option switched on, the CIPL would be based on the average speed and path throughout the test (excluding the parts of the test when the animal was actually in the zone).

Calculation of instantaneous speed

This option is only included to allow compatibility with results from experiments performed in versions of ANY-maze prior to V4.60. It should not be turned on for new experiments.

Prior to V4.60, the instantaneous speed reported by ANY-maze was the maximum speed, and this was a little exaggerated. A new method of calculation was introduced in V4.60, and we recommend that you use this new method unless you need to compare instantaneous speed results with experiments created in ANY-maze versions prior to V4.60.

Note that the new method is always used when making 'live' measurements for charts, so if this option is turned off, you may notice a difference between the values used on the chart when drawn in 'real-time' during the test, and when viewed post-test.

Removal of 'jumps' in the animal's track

This option is described in detail here.

ANY-maze help topic T0651