Voice Prompting Strategies for Educators

Use Natural Speech

Uses clear and natural speech. Natural intonation and pacing can make the interaction feel more like a conversation with a human. 

Use Simple and Clear Language 

Use simple, clear language, avoiding jargon and complex vocabulary  

Break Down Complex Requests 

If your request is complex, break it down into more straightforward, smaller commands.

Avoid Background Noise

The quality of the audio being received is an important measure of the accuracy of the results. Ensure a quiet environment to reduce misinterpretation of your commands.

Be Specific and Clear 

Use straightforward and easy-to-understand language. Avoid slang, jargon, or overly complex sentences.

Provide as much detail as necessary to convey your request accurately.

Use Respect and Sensitivity

Use polite and respectful language. Avoid commands that could be perceived as rude or demanding. The algorithm being used could model results on the user's input.

Be aware of cultural differences and sensitivities. Avoid language or references that could be offensive to different groups.

Ensure Accuracy and Truthfulness

Verify the information you provide or ask for is accurate and truthful.

Do not ask the assistant to speculate or provide opinions on sensitive or controversial topics.

Think About Ethical Usage

Avoid sharing sensitive personal information. Be aware of privacy implications when asking for or giving out information.

Ensure that your prompts do not lead to harmful actions or behaviours, either directly or indirectly.

Think About Inclusivity

Use language that is inclusive and avoids bias.

Ensure that your prompt is appropriate for the context in which the assistant is being used.

Conversational agents may need help understanding and interpreting user intent accurately, especially with ambiguous or complex queries.

Improved understanding of user intent is increasing with enhanced algorithms and training data.

User training is essential to understand what the limits are to the conversational agent 

Further detail in the prompts can help produce accurate and reliable information

Further practice can help a user understand how to frame questions properly and for the conversational agent to understand what the user is meaning

Users have different ways of phrasing queries and prompts; some may use slang, abbreviations, or non-standard grammar, which can be challenging for the agent to process.

User training is essential as the user will understand the software's limits.

Further use of the conversational agent will further understand what the user is trying to say.

Protecting user data and ensuring privacy is critical, especially in educational and personal applications.

Follow appropriate rules and regulations that are part of the operational policies.

Conversational agents can exhibit biases based on the data they are trained on, leading to unfair or discriminatory responses.

Regular audits of the material that the conversational agent creates are crucial in ensuring the system's reliability and trustworthiness. By conducting these audits, we can identify and address any potential biases or inaccuracies, further enhancing the quality of our service.

Give examples of the type of content you were expecting so it has a model to follow.