Enter your name, your company name, or any other name that helps identify the owner of this survey.
Type your last name, or your company name, or other name here that helps identifying the owner of this survey.
Tell LogRatio what type of survey this is by choosing one of four categories:
Select the language used for this survey
Enter your email address.
Attach the file with your survey data here.
Online survey software such as SurveyMonkey can export survey data in a CSV or XLSX file.
LogRatio is able to decipher the file structure of most online survey providers. Therefore, simply export your survey data, for instance in a CSV file and attach it here.
Click the Next button and LogRatio creates your professional survey report.
For purposes of trying out LogRatio, select your survey file (CSV or XLSX) or try one of the test files.
Select the date this survey started gathering interviews.
Select the date this survey stopped gathering interviews.
Tell LogRatio what to call this survey.
How large is the population this survey sample comes from?
For instance, if the survey is supposed to represent the whole Swiss population you may enter "8400000" in this field.
However, populations larger than one million are said to be "infinite". If the size of your population exceeds this figure simply type "1000000" in this field.
Sample surveys that back up important decisions should reproduce accurately, under probabilistic assumptions, the characteristics of interest of the surveyed population.
When launching new products, for example, we may desire accurate data, say with 95% Confidence Level. This means, if we repeat the same survey 100 times under the same conditions and with the same population, 95 times we shall get very similar results.
The confidence level of survey studies for business purposes is often set at 95%, but it is not unusual the case of levels ranging 90% to 99%. Note: the higher the confidence level the larger the sample and more expensive the survey.
Also called Hypothesis of the Study, this parameter can help significantly reduce the size of a sample without impact on the accuracy of survey results.
Suppose you want to estimate the proportion of respondents purchasing wine in the past month and you assume (better if you have evidence from a previous survey) that 35% did in fact buy wine. You can set the population proportion parameter to 0.35, this way reducing the sample size without negatively impacting the accuracy of the survey.
It requires 2’395 interviews to survey an infinite population at the 95% Confidence interval, 2% Error level, and Population proportion equal to 0.50. When the Population proportion is set to 0.35 the sample size reduces to 2’180 interviews.
If not sure, leave the Population proportion at 0.50 (or 50%) - indicating you do not have any prior information about the population under investigation.
LogRatio reports include five standard analyses:
LogRatio’s proprietary algorithm EAR (Entropy Auto-Recovery) understands, classifies, and transforms the answers to open-ended questions in a completely automated manner.
Traditional market research agencies do this process of classifying text by hand, and coding several hundred textual answers may take a lot of time. This is one of the reasons why surveys conducted by marketing research agencies are generally quite expensive.
At LogRatio instead we are developing technologies from natural language processing and machine learning that may change forever the way we analyze open text from survey answers. The algorithm is not perfect yet, we know, but we are working hard to make it better.
We recommend having at least 30 answers for one open question in order for LogRatio to deliver acceptable results. More is better.
Sentiment Analysis is about understanding and classifying the meaning of text.
LogRatio applies a proprietary sentiment analysis algorithm based on NLP (Natural Language Processing) and text analysis that interprets the meaning of answers, for instance to open-ended questions of a survey, and classifies them as Positive, Neutral, or Negative.
The whole process is automated and runs without human intervention.