est: Estimate Total Harvest

In peterdonati/hhss: Hunter Harvest Survey Simulations

Description

A generic function for estimating total harvest from objects created by the survey simulation functions.

Usage

est(simdat)

Arguments

simdat

Survey simulation data. Must be an object created from one of the three survey functions.

Details

In simulations with follow up surveys, two separate estimates are made. One estimate from the respondents to the initial survey and the other from respondents to follow up surveys. In the case of estimates from simple() outputs, the proportion of init_resp/init_sample is used to scale the initial estimate (i.e. the prop. of total respondents to total sampled is assumed to make up the same prop. of the entire population). Then, the follow up respondents are assumed to be representative of the remaining proportion of the entire population. This remaining proportion is used to scale the follow up estimate. Both scaled estimates are then added together to create the combined total population estimate.

This is similar to estimates made from census() outputs, except in that case the initial scaling proportion is init_resp/N.

Estimates for mand() outputs assume 100% reporting by successful hunters for initial reports if there is no follow up. If a follow up survey was simulated, it creates a harvest estimate from the follow up sample to estimate total harvest for the non-reporting portion of the population only, and then adds that to the sum of initially reported harvests.

Value

A data frame, containing the following variables:

• N: Hunter population size.

• resp_bias: Bias of successful hunters to report, relative to unsuccessful hunters.

• resp_rate: Underlying response probability, before any scaling/bias.

• mean_init_n: Average total initial survey responses.

• mean_fol_n: Average total follow-up survey responses.

• true_harvest: True harvest of population.

• min_hvst_est: Minimum harvest estimate of the survey repetitions (if any) at the response bias and response rate shown for that row.

• max_hvst_est: Maximum harvest estimate.

• mean_hvst_est: Average harvest estimate.

• mean_SE: Average standard error across repetitions.

• MARE: Mean absolute relative error.

• RRMSE: Relative root mean squared error.

See Also

pop and survey for how to create the data to be input to this function.

Examples

# First, make a population:
my_pop <- pop(N = 10000, split = 0.7, success1 = 0.25, success2 = 0.32)

# Then simulate a survey for that population:
my_pop_simple_followup <- simple(
  x = my_pop,
  sample = 0.5,
  resp = c(0.4, 0.6),
  bias = seq(1, 1.3, 0.1),
  fol = TRUE,
  fol_scale = 0.7,
  times = 10
)

# Finally, make your estimates:
est(my_pop_simple_followup)

peterdonati/hhss documentation built on Dec. 22, 2021, 7:45 a.m.