Description
In the web presentation, the Mechanical Reasoning test was introduced to twins as an activity called "Reinventing the wheel", with the following text: "You made it just in time! We must make sure never to find ourselves in this situation again. We need more builders! Tell me, are the rumours true? Did you really spend a year in the famous Dwarven School of Machinery? We have selected from our garrisons the brightest young minds who we feel are wasted learning only the art of war when there is building to be done. Will you take it upon yourself to set up a modest workshop and teach the young apprentices a thing or two?" This text was accompanied by the image below:
The test itself consisted of a practice item followed by 16 scored test items. Each item took the form of a conventional multiple choice question, comprising a stimulus image, a question, and 3 response options. Twins responded by clicking on one of the response options. On clicking a response, the response was immediately submitted to the web server, and the next item was presented. A full table of test items, including questions and stimulus images, can be found below under the heading Table of Test Items.
The test was preceded by an instructions page, which is reproduced here:
The instructions page was followed by a single practice item. The practice
item was optional: a twin could skip the practice item, attempt it once, or
attempt it many times if they wished. On each attempt of the practice item, the
twin was presented with feedback to say whether they chose the correct response,
and what the correct response should have been. No such feedback was given after
the main test items. No data for the practice item is included in the dataset.
The practice item is reproduced here with an example response, illustrating the
nature of the items in general:
The 16 main test items followed immediately after the practice item. The items were presented in numerical item number order, without any branching or discontinue rule. Each item had a 25 second time limit, and a count-down timer was shown at the top-left corner of the page for every item, as a reminder of this.
Twins were encouraged to complete the test in one session if possible. Hence, no "come back later" button was placed on screen at any stage. However, it was possible to halt a test, by closing the browser or turning off the computer. After halting the test, it was possible for a twin to resume at a later time or date by logging in again: they would be presented with the next item to be attempted.
An individual test item could be compromised by an interruption after the item had started and before it had ended. Interruptions might result from twin actions such as closing the browser, or clicking the browser's back or refresh buttons. Interruptions might also be caused by random events such as computer crash or loss of internet connection. Such events seem to have been rare. The effect of interruption was to forfeit the item with zero score. On logging in at a later time after an interrupted item, a twin would be presented with the following item of the test.
Web Test Rules
- Scoring
For each of the 16 items, the correct response is given a score of 1, and any of the incorrect responses is given a score of 0. The practice item does not contribute to the scoring. The maximum total score for the test is 16. - Timeout rule
A timeout of 25 seconds applies to each item. After 25 seconds with no response, the item is forfeited with zero score, and a "next item" screen appears. It was possible to pause indefinitely on the "next item" screen before resuming at the next item in the test.
Dataset Item Variables
The item variables relating to this test, as listed in the table below, were generated automatically by programs on the web server. For the sake of completeness, the table also lists derived variables relating directly to this test (rckmrdata1/2, rckmrstat1/2, rckmrtime1/2, rckmrtmn1/2). Note that the dataset only includes data for completed tests. If a twin started the test but left it unfinished, then the data for that test were deleted from the dataset.
Note that the item response variables have been recoded where appropriate with values -1 and -4 to denote item timeout or interrupted respectively. See the web data cleaning page for further details. The start and end dates/times have not been retained in the dataset.
Variables | Explanation | Values |
---|---|---|
rckmrdata1/2 | Data flag: are test data present? [DERIVED VARIABLE] | 0=no, 1=yes |
rckmrstat1/2 | Test status: was the test completed? [DERIVED VARIABLE] | 0=not started, 1=started but not finished, 2=finished successfully, 3=compromised by loss of data, 4=compromised by random responses |
rckmrsttm1/2 | Start date and time of the test [not in dataset] | Date-time values |
rckmrentm1/2 | End date and time of the test [not in dataset] | Date-time values |
rckmrtime1/2 | Time interval from start to end of the test [DERIVED VARIABLE] | Minutes (decimal values) |
rckmrtmn1/2 | Mean item response time [DERIVED VARIABLE] | Seconds (decimal values) |
rckmrXXan1/2 | Test item response for item XX* | 1, 2, 3 or 4 (denoting the response selected), -1=timed out, -4=item interrupted |
rckmrXXsc1/2 | Test item score for item XX* | 0=incorrect, 1=correct |
rckmrXXrt1/2 | Test item response time for item XX* | Milliseconds (integer values) |
rckmrtot1/2 | Test total score: total score for all 16 items | 0 to 16 (integer values) |
rckmrtotp1/2 | Test total score expressed as a percentage | 0 to 100 (decimal values) |
* XX denotes the item number, 01 to 16
Table of Test Items
The test items are tabulated below, for the practice item and the 16 main test items. In each case, the item number, stimulus image, stimulus text (question), response options and correct response are shown.