- #1

Soaring Crane

- 469

- 0

Thanks for any reply.

You are using an out of date browser. It may not display this or other websites correctly.

You should upgrade or use an alternative browser.

You should upgrade or use an alternative browser.

- Thread starter Soaring Crane
- Start date

In summary, the conversation discusses the use of binomial distributions and how to determine the best estimator for the parameter p or the mean and variance of the distribution. Different technical terms, such as "maximum likelihood" and "unbiased", are mentioned as possible ways to define the "best" estimator. Clarification and specific questions are needed in order to provide a mathematical answer.

- #1

Soaring Crane

- 469

- 0

Thanks for any reply.

Physics news on Phys.org

- #2

chiro

Science Advisor

- 4,817

- 134

What exactly are you trying to do? Are you trying to see if a distribution is binomial? Are you assuming its binomial to estimate its parameter?

Binomials are good for modelling sums of I.I.D Bernoulli (Yes/No, On/Off etc) type stochastic processes (i.e. random processes).

- #3

Soaring Crane

- 469

- 0

- #4

Stephen Tashi

Science Advisor

- 7,861

- 1,600

Soaring Crane said:which variability value (interquartile range, variance, standard deviation, etc.) are most appropriate for the data?

I think what you are asking is:

Given the data, what is the best method for estimating the parameter of the binomial distribution that fits it?

This is not a precise mathematical question until you define what "best" (or "appropriate") means.

In mathematical terms, you are seeking an "estimator" (i.e. an algorithm or formula whose input is the sample data and whose output is an estimate of the parameter p (the probability of "succsess") that defines a binomial distribution.) An estimator depends on the random values in a sample so the estimator is a random variable. Thus there is no guarantee the estimator will always be close to value you want to estimate. To say what a "best" estimator is, you must be specific about "best" means in scenario that involves random outcomes. Some common ways of expressing human tastes for "best" estimators have the technical names: "maximum liklihood", "unbiased", "minimum variance", "consistent".

So, to ask a question that has a mathematical answer, you should ask questions like:

"What is a maximum liklihood estimator for the parameter p of a binomial distribution?"

"Is there an estimator of the paramater p of a binomial distribuion that is both unbiased and has minimum variance?"

Some forum member can answer those, or you can find the answers on the web, now that the right jargon is established.

Edit: Or perhaps you don't care about p, but only about the mean and variance of the binomial distribution. In that case you should ask for estimators of those parameters. (Even though the mean and variance are both functions of p, a "best" estimator of p is not necessarily a "best" estimator of the mean or variance.)

- #5

blue_raver22

- 2,250

- 0

The choice of central tendency and variability values for binomial distributions depends on the specific characteristics of the data and the research question being addressed. Generally, the mean is the most commonly used measure of central tendency for binomial distributions as it takes into account all data points and provides a good overall representation of the data. However, if the data is skewed or contains outliers, the median may be a more appropriate measure of central tendency as it is less affected by extreme values.

In terms of variability, the interquartile range (IQR) may be a better measure if the data is not normally distributed or contains outliers, as it is less sensitive to extreme values compared to measures such as the standard deviation or variance. However, if the data is normally distributed, the standard deviation or variance may be more appropriate measures of variability as they take into account all data points and provide a more precise measure of spread.

Ultimately, the choice of central tendency and variability measures for binomial distributions should be based on a thorough understanding of the data and the research question, and multiple measures should be considered to provide a comprehensive understanding of the data. It is also important to consider the limitations and assumptions of each measure when making a decision.

The binomial distribution is a probability distribution used to model the probability of a certain number of successes in a fixed number of independent trials. It is commonly used in science to analyze the results of experiments and to make predictions about the likelihood of certain outcomes.

To calculate the probability of a specific number of successes, you can use the formula P(X = k) = nCk * pk * (1-p)n-k, where n is the number of trials, k is the number of successes, and p is the probability of success in each trial. nCk is the combination formula, which is used to calculate the number of ways to choose k objects from a group of n objects.

The values of n and p depend on the specific situation and experiment being studied. Generally, n should be a fixed number of trials and p should be a probability of success that is consistent across all trials. For example, if you are flipping a coin 10 times, n would be 10 and p would be 0.5 (assuming a fair coin).

No, the binomial distribution is used for discrete data, meaning data that can only take on certain values (such as whole numbers). It cannot be used for continuous data, which can take on any value within a range.

The binomial distribution can be seen in many scientific experiments, such as coin flipping, genetics, and drug trials. For example, in genetics, the binomial distribution can be used to calculate the probability of a certain number of offspring inheriting a specific trait. In drug trials, the binomial distribution can be used to determine the likelihood of a certain number of patients experiencing side effects from a new medication.

- Replies
- 3

- Views
- 803

- Replies
- 4

- Views
- 1K

- Replies
- 14

- Views
- 907

- Replies
- 0

- Views
- 1K

- Replies
- 14

- Views
- 6K

- Replies
- 4

- Views
- 2K

- Replies
- 1

- Views
- 1K

- Replies
- 8

- Views
- 2K

- Replies
- 28

- Views
- 3K

- Replies
- 5

- Views
- 2K

Share: