Sphinx – French company. The name means sphinx.

The company has been active for 35 years. They are advertised as a set of tools to “assess customer satisfaction, test a new product, or assess a new market. In addition, our software is perfectly suited to the needs of the human resources department to measure employee satisfaction, the work environment, training evaluation, and any ad hoc study.” Therefore, they focus on the creation of surveys, their management, through the creation of distribution lists or their own panels that allow monitoring; In addition to offering tools to analyze the data collected, in a qualitative and quantitative way, including, what would be its strong point, the communication of the results through reports, dashboard, stories or very attractive infographics. The accesses to the results can be customized according to the role of the person.

After analyzing their website, promotional videos and YouTube channel, we see that they advertise as a company to create and analyze data, erroneous or false claim. The surveys, even though they do allow the use of many graphic and design resources, do not allow the necessary recoding and previous debugging of more specific tools. The DATAVIV, the data analysis part, allows, as well as the Create surveys part, very interesting graphical and formatting features, as well as the use of dynamic filters to quickly segment any of the crosses or tables requested. But at a statistical level, what it shows is the approximate value of significance of bivariate crosses, not the test values or the intervals of the effects. The graph shows a cross between department (2) and general evaluation of the course, this would be a t-test, in case it is a parametric test -an aspect that Sphinx does not even raise and is essential-, the output shows an approximate degree of significance, an inter and intra variance that would not add anything to a research report or paper…among others, two key elements are missing for them to be useful results: the test value and the confidence interval of the difference between the means, without this the bivariate test and results are “lame”.

In advanced analyses, also very focused on visual and design aspects, essential tests are missing, such as logistic regression, such as reliability analysis – a part that is explained in courses on prospective, test construction and market research.

Its claim to be the ideal tool for researchers, for TFM, thesis... is short and limited to simple and descriptive works, both with survey methods and open questions.

In terms of surveys, it offers two options: DECLIC and IQ –more professional. Interesting in that, through its intuitive and very Windows-style menus, it allows formulating questions of very different formats and scales, both single and multiple answers. The most interesting thing, without a doubt, are the customization possibilities, at the visual and design, also the use of filters – much easier to apply and combine than in other free tools like google forms-, the possibility of grouping questions and retouching answers... All this has the same limitations as other tools, not allowing the answers to be coded for later statistical analysis, such as indicating that it gives us a 1 for females and a 0 for males , rather than the response's textual label which we should then replace or change from Excel.

The Analysis part, DATAVIV, allows you to generate very visual reports and in different formats, pdf among others. Allows you to customize the outputs, graphs, tables. It allows making crosses in a simple way, by dropping a second variable on the interface and filter option. Open responses are also graphed in a similar way, with different graphing options - most typically the word cloud, as in the Mentimeter tool. It allows graphing open responses as a factorial or dimensional analysis, it has options from which to correct misspelled words and assign them to the main words, and it allows, as with the closed survey variables, to apply filters to one output or to the entire survey. report.

The reports can be shared with other people, clients, collaborators... from the My views option you can change their visualization –to infographics, to dashboard… It has multi-platform options, a Language button from which the questionnaire can be translated into several languages, a URL can be generated to spread it in a mailing –in the Broadcast option.

In the Diffusion and Collection option, in the main menu, you can manage the data collection. The tool allows you to import Excel from emails, in addition to scheduling the sending of a survey and its follow-up – reminders.

In short, the tools it offers to create, analyze and communicate data are very visual, what stands out the most and differs from other options is that it integrates creation, analysis –even if it is very descriptive and unscientific- and the presentation of results. Very interesting for novice researchers who do not have to do exhaustive analysis or research, and also for educational use due to its ease of use, visual and design options.

Sphinx has a YouTube channel (Le Sphinx Iberoamerica) for Spanish speakers, where you can see applications and use of Sphinx.

One of the videos of the person in charge of Sphinx ends by saying “you don't have to be a statistics teacher to use this software”. And I think you are right, the use is intuitive and easy, but it is not useful for carrying out research or complex analysis, nor would it serve as a software reference for a paper or scientific article, since it is not a soft statistical.

My personal assessment, as a teacher, author of various materials, research and data modeling technician, in addition to being curious and an analyst of any soft or application that helps researchers, is that it is a interesting tool to generate non-professional surveys, with many options to design, launch, manage and make a panel of consumers, as well as to describe descriptive results with outputs very attractive and with the option apply dynamic filters that allow to intuit, draw, the incidence of second variables in the behavior of a univariate or bivariate distribution.

interest is more pedagogical, can allow -because of its attractiveness and functionality- students are more motivated to use the survey technique, also in the preliminary analysis it means tabulate and represent descriptive results. But it is completely insufficient to replace the use of soft such as SPSS or PSPP, since it does not offer the level of detail and results that statistics courses require. If the price for the university context is moderate, it could be an interesting complement.