Readme

This readme describes the FTR questionnaire, as well as providing a guide to how to recreate the methods reported in Robertson & Roberts (2020) and Robertson, Roberts, Dunbar, & Majid (2020). Researchers interested in using the FTR questionnaire should consult these publications, where it is fully described. The description in this repository is designed to be useful in conjunction with these publications.

brief description

The FTR questionnaire is a language elicitation tool, designed to elicit speakers use of future time referring language. It may be used for a variety of purposes, including linguistic typology and psychological, and behavioural economics experiments.

It is designed to elicit usage across a range of: temporal distances from speaker "now", speaker certainty (numerically represented in percentages, as well as categorically, i.e. "certain","neutral","uncertain"), and FTR modes, i.e. whether the context involved a prediction, an intention, or a schedule.

notes

The version of the FTR questionnaire which we recommend potentially interested researchers to use is given in ftr_questionnaire/questionnaire_master.xlsx and ftr_questionnaire/questionnaire_long.xlsx. These files contain identical data. The only difference is questionnaire_master.xlsx is in wide format while questionnaire_long.xlsx is in long format. The file code/questions_wide_to_long.py builds the long format from the master. Researchers seeking to change anything should therefore edit questionnaire_master.xlsx and then use questions_wide_to_long.py to build a new long file.

The script code/write_FTR_block.py is used to create the files in qualtrics_question_blocks/, which can be uploaded to the qualtrics survey hosting platform to create a usable survey block. This creates a usable block of questions which are formatted like this:

old and legacy versions

Older versions of the FTR elicitation questionnaire are given in ftr_questionnaire/drafts_e1_to_e4/questions.xlsx and ftr_questionnaire/drafts_e1_to_e4/questions_e3.xlsx. These files contain versions of the FTR questionnaire which were used to generate the data presented in Robertson & Roberts (2020) and Robertson, Roberts, Dunbar, & Majid (2020). A guide to how the items relate to the the data presented therein is given in the ftr_questionnaire/drafts_e1_to_e4/README.md in. It is not recommended researchers use these draft versions, as they contain significant inbalances between independent variables. They are included here for historical and transparency purposes.

FTR quesionnaire guide

The remainder of this readme will focus on explaining the significance of the columns in the recommended version of the FTR questionnaire. With reference to ftr_questionnaire/questionnaire_long.xlsx, the columns denote the following:

1. frame_distance: Denotes how far into the future from speaker "now" is explicitly stated or implied by each context.
2. first_person: Only relevant when ftr_mode == intention OR scheduling, indexes whether the context is in the first, second, or third person.
3. keep: Constant, left over from the writing process.
4. ftr_mode: Indexes whether the context is a prediction, intention, or schedule (see: Dahl, 2000).
5. rank: A subjective ranking of which questions the researchers thought were best (within each frame_distance.
6. notes: Constant, left over from the writing process.
7. copy_edited: Constant, left over from the writing process.
8. old_uneek: A key which is designed to relate updated questions back to earlier versions of the FTR questionnaire. Should not be used, as many questions may have been altered in the updating process even though they share the same old_uneek key as in previous FTR questionnaire versions.
9. word: The word which is used to denote the "certainty information" given to participants. Is usually certain/zeker, but is decided when ftr_mode == intention.
10. question: The actual quesion. Context is given in the [square brackets], target sentence afterwards.
11. uneek a question index number which is unique across language, and certainty, conditions. That is to say, questions with the same question_form but which are altered to acommodate different certainty conditions will have different uneek indexes. Likewise, the same question given in English and Dutch will have different values of uneek.
12. certainty: Numerical representation of "certainty information" designed to be given in percentages. Always 100 in the certain condition, one of [40,50,60] in the uncertain condition, and not given [-999] in the neutral condition.
13. question_form: Indexes the general form of the question, i.e. is not unique across language, and certainty, conditions. That is to say, which are generally similar but are altered to acommodate different certainty conditions will have the same value of question_form. Likewise, the same question given in English and Dutch will have the same value of question_form.
14. language: Indexes the language of the question (english or dutch).
15. probability: qualitative categorical representation of the certainty condition. In the neutral condition no certainty information is given, in the certain condition, participants are informed they should imagine they are 100% certain, and in the uncertain condition, participants are informed should imagine they are one of [40,50,60]% certain. Since there is one intention and scheduling question per frame_distance, these are given 50% certainty in the uncertain condition. Since there are 3 prediction questions per frame_distance these are given one each at 40%,50%,60% certainty.
16. question_no: Index which is unique across levels of probability but not unique across levels of language, i.e. within a level of question_form, a question which had been altered to acommdate different levels of certainty information would have different value of question_no, but value would be the same across translated versions of the same question.
17. question_eng: An English translation of each question (the same as question when language == english).

`