Reliability of experiment and of results (1 Viewer)

NexusRich

Member
Joined
Jul 6, 2019
Messages
87
Gender
Male
HSC
2021
My teacher said that the reliability of experiment and of results are two different things and that repetition of trials will not improve the reliability of the experiment, could anyone pls explain the difference and how to improve the reliability of both ?
 

jimmysmith560

Le Phénix Trilingue
Moderator
Joined
Aug 22, 2019
Messages
4,116
Location
Krak des Chevaliers
Gender
Male
HSC
2019
Uni Grad
2022
I think that, since we know for sure that reliability of results refers to how consistent our results are after each trial, maybe your teacher was referring to the nature of the experiment when they mentioned that repetition of trials will not improve the reliability of the experiment.

For example, using an incorrect/irrelevant method as part of an experiment in an attempt to determine the nature/value of something renders the experiment itself unreliable, even if you can achieve consistent results using such a method after multiple trials.

That's what I think 🤔
 

Leadmen4y

Well-Known Member
Joined
Feb 26, 2021
Messages
477
Gender
Male
HSC
2021
A reliable experiment has minimal chance of random errors, which are fluctuations of the data as a result of instability of environment, these errors are often irreproducible (contrasting to systematic errors). That is my interpretation of what your teacher meant.

Random errors can be reduced by:
  • Ensure measuring instruments are easy to read.
  • Ensure measuring instruments are calibrated.
  • Conducting the experiment within a stable environment.
Lastly, repeating an experiment DOES NOT make it reliable. Instead, it gives you the opportunity to determine whether it is reliable or not through ascertaining the extent of consistency of the data.
 

idkkdi

Well-Known Member
Joined
Aug 2, 2019
Messages
2,454
Gender
Male
HSC
2021
Calibrating is more so a systematic error.

just to add on, i reckon HSC does a bad job with reliability. They seem to basically mix in precision with reliability, or at least the way i've seen it been written about.

but ye, for hsc, treat consistency of data per measurement and per experiment both as reliability.

an example of reliability per experiment, is, do the experiment again with a different batch of the same chemicals, and check if the results from the two experiments match up.
 

Leadmen4y

Well-Known Member
Joined
Feb 26, 2021
Messages
477
Gender
Male
HSC
2021
Calibrating is more so a systematic error.

just to add on, i reckon HSC does a bad job with reliability. They seem to basically mix in precision with reliability, or at least the way i've seen it been written about.

but ye, for hsc, treat consistency of data per measurement and per experiment both as reliability.

an example of reliability per experiment, is, do the experiment again with a different batch of the same chemicals, and check if the results from the two experiments match up.
Good point, definitely more of a systematic error. And yes, my teacher confirmed that precision and reliability are usually interchangeable.
 

Eagle Mum

Well-Known Member
Joined
Nov 9, 2020
Messages
532
Gender
Female
HSC
N/A
My teacher said that the reliability of experiment and of results are two different things and that repetition of trials will not improve the reliability of the experiment, could anyone pls explain the difference and how to improve the reliability of both ?
Repeating measurements during the same trial will enable reliability of measurements (ie. precision) to be estimated.
Repeating measurements in repeated trials will enable reliability of experimental design to be assessed, but doesn’t improve reliability unless the factors that cause variation(s) between trials are identified and controlled.

A simple example would be an experiment that sets out to determine how a catalyst affects the rate of a reaction. If the ambient temperature also affects reaction rate, then repeating the experiment throughout the day (eg. different classes perform the experiment during cold mornings vs warm afternoons) might produce different results which reflect on the reliability of the experimental design, whereas multiple groups within the same class each taking measurements of the reaction, with and without the catalyst, during the same experiment trial would address reliability of measurement.
 
Last edited:

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top