Frequently asked questions - answered in a dialogue

Q: I heard a lot of readability formulas, and there are a lot of websites that offer readability analyses, some of them are free. Why should I buy TextQuest/Refo?
A: There are many reasons: It provides results for more languages than any other package, it has builtin lists for known words, exception lists for syllable counting, abbrevations lists for several languages, and special features or handling headers and enumeration lists.
Q: Impressing. What about languages, can you name a few?
A: English, French, Spanish, German, Scandinavian languages, Slavik languages, and Turkish, in total nearly 100 different readability formulas.
Q: That's a lot. What are other advantages?

A: Well, all readability formulas require information on the language, the text genre, and the readership of a text. Most readability formulas are for English and any text and no special readership, but others like Dale-Chall are for children.
Q: So I cannot choose any formula?
A: Yes, you must know your text. If you use TextQuest/ReFo, you don't have to care about this: you specify the language, the genre, and the readership of a text and the suitable readability formulas will be used for the analysis.
For example if you use a readability formula for English text on a text in another language, the results are not valid. Or: English readability formulas applied on a text in German will show that German is difficult.
Only the readability formulas that are appropriate for your text are selected.
Q: Good to know, what about other problems?

A: TextQuest/ReFo provides precise results, other software doesn't.
Q: The results are not correct?
A: Indeed, and the reasons are manyfold.
Q: That's hard to believe. Please explain why.

Syllable counting in English

A: Most of the readability formulas count the length of words in syllables. In English this is difficult, because there are big differences between spoken and written language.
Q: Well, counting syllables is easy. You speak and count.
A: The text you want to analyse is not spoken, it is written as a sequence of letters.
Q: You count the diphtongs and vowels, and that's it.
A: What about words like ride, bike, smile? They have two vowels but only one syllable.
Q: Hmm - the e at the end is silent - you just ignore the e at the end of a word.
A: What about words like wire, hire, fire?
Q: The e is not silent if the letter r is front of it.
A: What about cafe?
Q: You can make a list of exceptions.
A: That is a lot of work, and the silent e is not the only problem.
Q: So what is the solution that TextQuest/Refo uses then?
A: It uses the work of the Carnegie-Mellon University. It provides a dictionary of more than 100,000 English words and their syllables. TextQuest/ReFo uses this dictionary to count the syllables.
Q: Hmm, and what happens if a word of my text is not in this dictionary?
A: That is very unlikely and did not happen during the tests. But if this is the case, a simple algorithm will count or better estimate the number of syllables.
Q: And can the effects be ignored?
A: That depends on your text. If you have huge amount of texts, certainly not.
Q: What formulas are effected?
A: Many readability formulas use the counting of syllables, some only count words with one syllable, others count long words with three or more syllables - that depends on the formula. You want a list of these formulas? Well: Flesch's REI from 1948, all derivates from Farr, Jenkins, Patterson and Pearl, Sumner, Kearl, Gunning's FOG index, Coleman's formulas, Fry's table from 1968, Forcast 1973, Linsear Write/Fog Count, McLaughlin's SMOG, and Wheeler/Smith.
Q: These are the most popular and known ones.
A: You also should know that many words are exceptions, so we added exception lists for the languages that require syllable counting. This was a lot of work, oftne more than 1,000 words for a language.

Wordlists

Q: You mentioned more problems, so what else is unique to TextQuest/ReFo?
A: The wordlists that are used in some formulas like Dale-Chall, Spache, or for German Bamberger/Vanecek's formulas.
Q: This is easy, you can download the lists from the net, and that's it.
A: This is not the case, because the authors of the formulas provide detailed information how and what to count. For example the lists that are available on the net only include the base forms of nouns, verbs, adjectives, and adverbs, but the authors write that you must add all forms, for example the plural forms for nouns, the past tense and past participle for verbs etc.
Q: So TextQuest/ReFo has all these words?
A: Yes, and so e.g. the Dale-Chall list has more than 8,000 entries, more than double of the original list.
Q: What about doing that myself?
A: You will see that this takes a lot of time, and many words in English can appear as a noun and verb, sometimes even as an adjective. We spend much time on it, so you don't have to reinvent the wheel.
Q: Anything else I should know?

Standardization

A: What about the size of your texts? Do you often have texts longer than 1,000 words?
Q: Yes, I often have very long texts, is that a problem?
A: Not at all, all formulas are standardized on text size.
Q: So I get the values of all selected formulas?
A: Of course, but that's not all. You'll get them for all paragraphs of the text also.
Q: But what is a paragraph?
A; You hit hit the return key twice so an empty line is in your text file. Or technically, CR/LF CR/LF.
Q: Wow, I think separate analyses by paragraph is a feature the free readability programs don't have.
A: Only one has, and the values for each paragraph show you the differences between the paragraphs, so there are hints which paragraphs might need improvement.
Q: Improvement, how?
A: Most readability formulas count the length of words and sentences. All long words and long sentences can be highlighted. Depending on the selected formula, the words in a word list can be highligted as well, e.g. Dale/Chall or Spaulding word lists.
Q: So all words from the lists are included?
A: Yes, and you can show each hit in your current file.
Q: If I have the results, can I analyze them statistically?

Results for the whole text and each paragraph

A: This feature is implemented, you can view the results for the whole text and each paragraph separately, and there is a table at the end of the log data.
Q: But what if I have a heading without a period.
A: If you make the heading a paragraph this will do. TextQuest/ReFo will treat the heading as a paragraph.

Enumerations

Q: Some of my texts include a lot of enumerations, this will make the sentences very long. If I understand you right, this will result that the text will be more difficult to read.
A: If your enumeration starts with a dash or a number, TextQuest/ReFo recognizes this and treats each enumeration as a single sentence, so that you don't have to insert periods yourself.
Q: This saves a lot of time and errorneous work.
A: That's why we implemented this feature.

Splitting a text into normal sentences

Q: You said that a text is separated into grammatical sentences. In most cases this is a period, but periods can also be used in numbers or abbrevations. How is this problem tackled?
A: Numbers are easy, because there is no blank after the number. Abbrevations are more difficult, because sometimes these consist of more than one letter. We collected frequent abbrevations in a list for many languages, so these abbrevations are detected and the periods after them are not a sentence end marker.
Q: But in English there are thousands of abbrevations.
A: In other languages as well, and it was a lot of work collecting all these from different dictionaries.

Features of TextQuest/ReFo

Q: All right, but how can I work with the software?
A: There is a text window, you copy your text into it, select language, text type and readership and press Text analysis. After a short time a windows open and shows you the results. If you click on a paragraph symbol, the results for this paragraph is shown.
Q: That's easy.
Q: How can I use the results in my text processor?
A: You can store it in the clipboard or as an external file.

Versions and licenses

A: You can export the data as a CSV file, so you can use this file for MS-Excel or statistical software like SPSS. If you did an analysis, just click on the export button, and the file will be copied to a destination you can specify.
Q: I only plan to use this software for my master thesis? Are there any discounts for me?
A: You can obtain a time limited license, e.g. for 3 months, and you benefit from licenses for academmic purposes - non commercial use.
Q: So I buy a 3 month license for academic use, right?
A: Yes. There are different versions, the more expensive ones include more languages and a vocabulary module.
Q: What does this module do?
A: It produces a word list, you can sort this by alphabet or frequency, ascending or descending. And you can remove stop words.
Q: Why should I do this?
A: Stop words are words that have a relative meaning only like articles, pronouns, preposionts, or conjunctions. These occur with a high frequency, and you can ignore these.
Q: And what else can I do?
A: You can check your text for typos, and a word list can also be a collocation list, e.g. which words occur simultaneously like United Kingdom, European Union, or the like.
Q: Can I test TextQuest/ReFo?
A: You can download a test version that allows you to work 30 days starting with the first execution. After that you can buy a license, enter the key, and you can continue your work.
Q: Are there any feature I cannot use, e.g. saving files or printing?
A: No, we want you as a satisfied customer. If you think our software is good, tell everyone else. If you detect errors or need new features, contact us.