In most published protocols, a choice is NOT recorded just because the insect crosses the junction.

👉 Standard practice:

• The insect must enter an arm and pass a defined decision point
• This is usually:
  • 1/3 to 1/2 of the arm length, OR
  • A pre-marked line (~3–5 cm from the junction)

✔️ So:

• ❌ Crossing the junction = NOT a valid choice
• ✔️ Crossing a decision line in one arm = VALID choice

👉 Why? Because insects often “probe” both arms briefly. Recording at the junction would give false positives.

✔️ Yes — this is usually considered a valid and final choice IF:

• The insect crosses the decision threshold (halfway or marked line)
• And stays oriented forward (not just touching and returning immediately)

👉 Many labs define:

• “Choice = insect moves X cm into one arm and remains for ≥ 5–10 seconds”

This is important 👇

Case A: Did NOT cross decision line

• ❌ Not counted
• Recorded as:
  • “No choice” OR
  • “Undecided”

Case B: Crossed decision line, then came back

• ✔️ Usually counted as a choice already made
• Movement back is ignored

👉 Reason: The first committed movement is considered the behavioral response.

There are two approaches, depending on study type:

A. Choice-based studies (most common)

• Only final choice is recorded
• Time is secondary:
  • “Time to first choice” (latency)

B. Behavioral analysis studies

Time is recorded as:

• Time spent in each arm
• Time in central zone
• Number of entries into each arm

👉 This is done by:

• Stopwatch (manual)
• OR video tracking software (preferred in research)

Typical duration:

• 3 to 5 minutes per insect

Rules:

• If insect makes a choice → stop early
• If no choice within time limit → mark as:
  • “No response” / “Non-responder”

👉 Important:

• Non-responders are usually excluded or reported separately

Typical:

• 20–50 insects per treatment

And:

• Repeat experiments across days for reliability

Good Y-tube experiments follow these controls:

✔️ Airflow control

• Equal airflow in both arms
• No turbulence

✔️ Odor switching

• Swap odor sides regularly (to avoid positional bias)

✔️ Cleaning

• Clean Y-tube after few insects
• Prevent odor contamination

✔️ Control test

• Run clean air vs clean air
• Expect ~50:50 distribution → confirms no bias

Most common output:

• % insects choosing odor A
• % insects choosing odor B
• % no response

Statistical tests:

• Chi-square test
• Binomial test

• Crossing junction ❌ = not counted
• Crossing halfway ✔️ = counted
• Returning after crossing ✔️ = still counted
• Time measured = optional (latency or duration)
• Observation time = ~3–5 min
• Non-responders = recorded separately