Over a million developers have joined DZone.

Survey Analysis for SAE Institute: A Case Study

DZone's Guide to

Survey Analysis for SAE Institute: A Case Study

This case study describes how ParallelDots carried out a detailed survey analysis to make sense of the general sentiment of students at the SAE Institute from Australia.

· AI Zone ·
Free Resource

Did you know that 50- 80% of your enterprise business processes can be automated with AssistEdge?  Identify processes, deploy bots and scale effortlessly with AssistEdge.

SAE Institute from Australia is considered the front-runner in creative studies. They recently ran a survey to understand the general sentiment of students and identify key areas of improvement. This case study describes how ParallelDots carried out a detailed survey analysis to make sense of the student responses and derived specific insights from this data. 750 students responded to the survey, resulting in a total of 4500 comments.

The Challenges in Traditional Survey Analysis

Traditional methods of survey analysis suffer from multiple human biases and are intensive in terms of effort as well as time. Knowing this, the institute decided to outsource the project to ParallelDots.

We employed our in-house tools to perform these tasks. Our standard text classification product is SmartReader.

Overview of the Resultssurvey analysis

Key Insights Generated as a Result of Survey Analysis

Insight 1: General Sentiment towards the institute

Research has shown repeatedly that allowing respondents to provide feedback unaided and in their own words can yield data that is more predictive of actual attitudes and behavior.

ParallelDots Emotion Analysis API tracks the primary six emotions expressed in verbatim comments that fundamentally drive human behavior: happiness, excitement, fear, sadness, anger, and boredom. Here’s a breakdown of the sentiment analysis of students’ comments.

survey analysis

Interestingly, most students were not only positive about their experience at the institute but also excited. Clearly, the student community at SAE Institute is enthused and finds positive value in their education. This is good news for the administration.

Insight 2: Major Areas of Concern as Revealed by Survey Analysis

This step involves using SmartReader’s built-in predictive analytics to identify, in this case, six topics that are significant predictors of satisfaction. These are:

  • Curriculum
  • Teaching
  • General Facilities
  • IT Facilities
  • Environment and
  • Administration

For a live study, more such themes can be added to the survey analysis.

The most popular topics of concern were teaching and general facilities. 59 percent of the overall responses pointed to these themes.

Delving into each of these themes, we also explored the net sentiment score.

Insight 3: Key Areas of Success

survey analysis

The curriculum stood out as a popularly liked aspect of the institute. Most students saw it as an addition, rather than a deterrent, to their learning. More importantly, the net sentiment regarding curriculum stands at 69 percent. This strongly indicates that the student community is content with the curriculum.

Digging into the comments that pertain to the theme of curriculum, a keyword cloud was generated.

survey analysis

Clearly, the students strongly believe that their learnings are adding to their skillset and enjoy the courses offered. The institute also offers a trimester-long internship, which appears to be popular and liked.

Insight 4: Key Areas of Improvement

Areas for improvement are environmental conditions and administration support. However, the negative net sentiment is 32 percent only.

survey analysis

The most popularly occurring keywords that appear in comments pertaining to the administration indicate that the students are dissatisfied with how frequently their classes are rescheduled and canceled.

Summary and Highlights

survey analysis

By leveraging responses to open-ended comment questions together with our AI-based solution predicted satisfaction from student comment data. The ParallelDots survey analysis provided much deeper and more actionable insights which cannot be obtained through traditional methods of survey analysis. In contrast to typical tracking studies, which are limited to just measuring satisfaction levels, SmartReader yields not only intelligence about the actual factors driving satisfaction but also quantifies the extent to which each factor actually influences satisfaction.

The institute found ParallelDots’ solution effective: It provided an automated way of turning hundreds of student voices into a clear understanding of what matters to different groups and find specific improvement suggestions. Some of these suggestions, like streamlining the timetable, were truly novel and actionable.

For those of you who are interested in knowing more about the technology that ParallelDots uses to carry out survey analysis, please read on.

About SmartReader

survey analysis

SmartReader is a simplified SaaS solution where you can do all this in one go! All you’ll have to do is upload your survey responses in Excel and wait for the AI to do its job. You can customize the themes, give your own keywords, test the results, and tweak the project until your results are good.

The Technology Used for Survey Analysis

Contextual Semantic Search (CSS)

CSS is an unsupervised technique that creates a reasonable accuracy text classifier with zero training examples. This is an AI-powered smart search algorithm and a unique offering by ParallelDots. It employs few-shot learning and requires very few training examples.

Smart Keyword Generator

The idea is to intelligently extract keywords and phrases that capture the underlying message of a sentence. This is a mixture of keyword extraction, entities, and part of speech identification (noun, adjectives, adverbs, verbs). This part of the survey analysis allows reveals the key areas of improvement.

Sentiment Analysis

The ParallelDots sentiment analysis algorithm can determine whether a statement is positive, negative, or neutral based on the use of language. It can be customized to become domain intelligent.

Emotion Analysis

Emotion analysis goes one step further and determines the variant of sentiment that a feedback statement expresses. It can classify text statements into happy, sad, angry, fear, excited, and bored.

Consuming AI in byte sized applications is the best way to transform digitally. #BuiltOnAI, EdgeVerve’s business application, provides you with everything you need to plug & play AI into your enterprise.  Learn more.

nlp apis ,nlp ,deep learning for ai ,machine learning algorithms ,artificial intelligence ,machine learning

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}