Survey Translation & Linguistic Coding: Get Qualitative Feedback And Quantifiable Data
You want to grow your business and make it more client-oriented. So, you need data! You think about conducting a survey, but the kind of feedback you are looking for will require customers to share their experience in their own words, not simply tick a box with an answer you have chosen for them. But here’s the rub, you have customers from all over the world and you will need all of their feedback.
So, what does effective survey translation look like? And how can you measure multilingual responses with open-ended questions effectively? Let’s talk about linguistic coding!
Why Do I Need Surveys To Grow My Business?
In order for a business to expand effectively, information and honest feedback are required. This data should be implemented in the areas of the business that need improving. Feedback gained from surveys will:
- Open you up to possible blind spots you have in your business strategy.
- Gain valuable insights into what your consumers truly think of your brand.
- Generate brand awareness and collect information from new markets that will help outline an international business model.
- Use both negative and positive feedback to create a more customized experience for different groups and markets.
Businesses that opt to evolve from a customer-oriented model perform better than those that don’t. In fact, research has shown that 54 % of US buyers say that most companies urgently need to improve their customer experience.1 Customer relations do not end when they have purchased the product. This is where surveys play a key role; they are a great tool that enables further interaction, especially when they are designed to give complete and honest feedback.
Survey Translation: How To Get Reliable Data From International Markets
According to data collected by Microsoft between 2015 and 2019, 89% of consumers surveyed across the United States, the United Kingdom, Germany, Brazil, and Japan, stated that organizations should provide consumers the opportunity to give feedback.2 The question is: would these consumers prefer to be asked and provide their opinions in their own language or a foreign one? Of course, you might be tempted to use English. However, that wouldn’t be ideal.
A report by EF Education First that ranks English proficiency around the world revealed that Northern Europe had the highest score, ranging from 60 to 80 out of 100, while the rest of the world showed mixed ranges (Moderate, Low, and Very Low).3
In other words, English might be widely spoken but often not as fluently as you’d expect. Just because someone speaks English doesn’t mean they can speak it well (or even feel confident enough to use it). Translating your survey into the languages of your audience then is the ideal scenario, which won’t only mean more accurate data but also higher customer satisfaction.
However, once you have decided you require survey translation, there are still several elements to consider in order to safeguard the quality of the survey translation. If the quality is lacking, not only will you get inaccurate feedback that will not offer you the data you need, but your consumers might not trust your brand if it fails to provide clear and accurate communications. So let’s dive right in and focus on some of the specifics.
Human Or Machine?
Survey questionnaires are all about collecting reliable data, which means that the translations should reflect an accurate version of the original. Screener, demographic, and survey questions are carefully crafted to extract the most accurate data from participants, which is why we strongly recommend sticking to human translation and foregoing machine translations.
Despite its rapid growth in recent years,4 machine translation still lacks the contextual awareness to make translation decisions that incorporate the cultural and linguistic awareness a human translator has. And even Machine Translation Post-Editing can fall short here, because the time that should be saved by running a source text through an MT-Engine, might be lost when the linguist has to make too many alterations.
For example, more than just a word-for-word translation, linguists will need to make an informed decision about which language form should be leveraged and which words should be used depending on the linguistic, regional, and cultural framework. For example, let’s have a quick look at Chinese for Survey Translation;
- Chinese has two written forms in traditional and simplified. Perhaps the survey will be directed at Mandarin speakers? Or immigrants? The linguist will have to look at the demographic to decide which form will be best suited.
- It is important to note that the conventions of an English language survey should also be adapted to the Chinese language. For example, Chinese characters are not written with the distinction between upper and lower case letters, nor are the characters alphabetical. These would require individual adaptations, even after MT.
That is not to say that automation doesn’t have a part to play in survey translation. Translation (CAT) Tools and Translation Memories (TM) will play a vital part in streamlining the process, and indeed Linguistic Coding can support the creation of a goal-oriented model. More on both of these below!
How Accuracy In Survey Translation Is Assured
So, let’s get down to the nuts and bolts and have a look at what the actual translation process would look like. The content will have to go through a TEP (Translation, Editing, Proofreading) Process. Qualified LSP will leverage Translation CAT Tools enriched with a glossary that includes product-specific terminology and a style guide that holds instructions specific to the brand. They will also make use of TM so that any previously saved terminology can be saved and inconsistencies prevented.
Since accuracy is of the utmost priority, only native, in-country linguists will be engaged in the translation projects and specific translation methods will be utilized. For survey translation, Back-Translation is commonly used so that inaccuracies in terminology, cultural ambiguities, and grammatical errors, can be weeded out.
Back Translation can be further divided into single-back-translation and double-back translation. Double Back Translation Service offers an extra level of security that will serve surveys very well, since they leave absolutely no room for ambiguity. During double-back-translation, the client offers the original text which is then translated. Subsequently, two independent translation teams translate the documents back to their source language, while a third Linguistic reviewer uses both translations and the notes provided by the clients, in order to create the most complete and accurate version.
Pro Tip! Survey Translation can include more elements such as a streamlined prepping stage and specific information to provide to your LSP. Learn more about it in our blog International Survey Translation: This Is How Accuracy Remains A Priority!
Do You Need To Process Open-Ended Questions? Enter Linguistic Coding!
In order to reap the benefits, surveys could provide your business, the flow of multilingual information needs to be measured consistently to maintain a goal-oriented model. And if you want to receive honest and open insights such as strategy demands, surveys featuring open-ended questions are often required.
This is where linguistic coding, also referred to as open question coding or open-ended coding, comes into the picture. Linguistic coding translates, organizes, compiles open-ended questions in a format that data analyzing software can easily process.
So, how should these open-ended questions be integrated? Not blindly, that’s for sure. In many situations, it may seem advisable to avoid using open-ended questions whenever possible, and, unless an in-depth response is required, stick to multiple-choice questions. But it all depends on your goals.
While leveraging linguistic coding can help quantify the responses, open-ended questions should still be used in a very strategic way when applied. Customers will gain the opportunity to share their experiences and feelings with no restraints, and you will get insights and new angles for your business you may not have thought of in the past. However, make sure you do not use these in excess. While open-ended questions can be valuable, they can also discourage people from taking or finishing the survey if the questions are too long-winded. So, make sure you find a balance and design the open-ended questions in a way that will get you the data you really need.
How Does Linguistic Coding Work?
Applying linguistic codes effectively will turn the challenge of linguistic nuances into a strength and convert subjective and qualitative feedback into quantifiable materials. Linguistic coding is a combination of human expertise and partial automation. This means that linguists, reliable tools, and data scientists will need to join forces in order to have the process go down without a hitch. And this process will include:
- Accurate translation of feedback by native, in-country experts
- The translated feedback will then be divided into specific quantitative criteria related to the questions and what the business wants to learn from the responses.
- The criteria are compiled and processed in a system so results can be collected.
For example, let’s assume one of the museum’s open-ended questions is Did you enjoy the exhibition?â€. A piece of translated feedback could be: It was pretty informative and the artworks were put together in a beautiful way and the pieces themselves were splendid.
A scale will be attributed to the underlined terms that will vary depending on the business’ preference. If we assume a conventional approach of 1 to 5 stars, Pretty informative will perhaps be given the metric of 3 stars, while beautiful could be graded as 4, and splendid 5.
In this way, adjectives can be used as a reference for the value of a certain review, and the linguistic coding can assemble this in a more quantifiable framework that is easily processed.
Conclusion
Survey Translation and Linguistic coding are very useful services for businesses that want to strengthen their client-oriented approach on an international level. When it comes to surveys, open-ended questions may seem like a real challenge to measure and process, but sometimes, you simply need open and unconstrained feedback from a customer in order to fully meet your expansion goals and keep clients happy. This is why open-ended questions should always be applied in a strategic way, and linguistic coding should only be performed by qualified linguists, specialized tools, and expert data scientists.
At Laoret, we believe that humans and tools can make each other better. That is why we hire only the most specialized, native, in-country linguists and ensure they possess full literacy in our custom-designed tools.